v3a bpg fra final fmmp 20131213ok...final report, volume 3a “best practice guidelines for flood...

Str

B

Flood

ructurain

Best Flo

Mekon

d Managem

al Meathe Lo

Practood R

ng River

ment and

asures ower M

tice GRisk A

Volum

May 2

r Comm

Mitigatio

and FMekon

GuideAsses

me 3A

2010

ission

n Program

lood Png Basi

elinessmen

mme

Proofinin

s for nt

ng

Published in Phnom Penh, Cambodia in September 2013 by the Mekong River Commission, Office of the Secretariat in Phnom Penh Citation: Royal Haskoning, Deltares, UNESCO‐IHE, The Flood Management and Mitigation Programme, ‘Component 2: Structural Measures and Flood Proofing in the Lower Mekong Basin’, May 2010, Final Report, Volume 3A “Best Practice Guidelines for Flood Risk Assessment”. 158 pp. Opinions and interpretations expressed are those of the authors and may not necessarily reflect the views of the MRC Member Countries Editors: Ms. Tiffany Hacker, Dr. David Lampert, Mr. David Smith Editors have applied, to the extent possible, the MRC standard for names of rivers, villages, districts and provinces. However some names in maps, figures and tables could not be timely adjusted as a result of the picture‐format used by the authors. © Mekong River Commission Office of the Secretariat in Phnom Penh (OSP) 576, National Road #2, Chak Angre Krom, P.O. Box 623, Phnom Penh, Cambodia Tel. (855‐23) 425 353. Fax (855‐23) 425 363 Office of the Secretariat in Vientiane (OSV) Office of the Chief Executive Officer 184 Fa Ngoum Road, P.O. Box 6101, Vientiane, Lao PDR Tel (856‐21) 263 263. Fax (856‐21) 263 264 Website: www.mrcmekong.org Email: [email protected]

MRC Flood Management and Mitigation Programme Component 2: Structural Measures and Flood Proofing

Guidelines Flood Risk Assessment ‐ i ‐ May 2010

TABLE OF CONTENTS

1 INTRODUCTION .................................................................................................................... 3 1.1 Guide to the reporting structure of the Flood Management and Mitigation

Programme ‐ Component 2, Structural Measures and Flood Proofing .......................... 3 1.2 Best Practice Guidelines for Integrated Flood Risk Management .................................. 4 1.3 Concepts of Flood Risk Assessment in the Best Practice Guidelines .............................. 4

1.3.1 Flood risk and flood risk management .......................................................... 4 1.3.2 Flood risk assessment .................................................................................... 5

1.4 Basis of the Flood Risk Assessment Guidelines .............................................................. 7 1.5 Use of the Flood Risk Assessment Guidelines ................................................................ 7 1.6 The Best Practice Guidelines and project phases/stages ............................................... 8

2 FLOOD HAZARD ASSESSMENT ............................................................................................ 11 2.1 Introduction .................................................................................................................. 11 2.2 Classification of type of flood ....................................................................................... 11 2.3 Tributary flood hazard assessment .............................................................................. 15

2.3.1 General ........................................................................................................ 15 2.3.2 Flood hazard derived from rainfall extremes .............................................. 15 2.3.3 Example of flood hazard derived from rainfall extremes ............................ 18 2.3.4 Flood hazard determined from observed flows .......................................... 23 2.3.5 Flood hazard derived from regional flood statistics .................................... 26

2.4 Mainstream flood hazard assessment.......................................................................... 28 2.4.1 General ........................................................................................................ 28 2.4.2 Data collection ............................................................................................. 28 2.4.3 Database development and conduct of field visits ..................................... 30 2.4.4 Data validation and processing ................................................................... 31 2.4.5 Hydrological hazard assessment in terms of flood peak water levels

and flood volumes ....................................................................................... 31 2.4.6 Flood hazard assessment ............................................................................. 32 2.4.7 Example: Mekong near Nakhon Phanom (Thailand) ................................... 33

2.5 Combined flood hazard assessment ............................................................................. 37 2.5.1 General ........................................................................................................ 37 2.5.2 Data collection ............................................................................................. 38 2.5.3 Database development and conduct of field visits ..................................... 39 2.5.4 Data validation and processing ................................................................... 39 2.5.5 Hydrological hazard assessment ................................................................. 39 2.5.6 Flood hazard assessment using a hydraulic model ..................................... 41 2.5.7 Example: combined floods in the lower Xe Bang Fai River Focal Area

(Lao PDR) ..................................................................................................... 42 2.6 Flood Hazard Assessment in the Mekong Delta ........................................................... 51

2.6.1 General ........................................................................................................ 51 2.6.2 Data collection and conduct of field visits ................................................... 51 2.6.3 Database development ............................................................................... 53 2.6.4 Hydrodynamic model development ............................................................ 54 2.6.5 Flood hazard assessment ............................................................................. 58

2.7 Flood hazard mapping .................................................................................................. 58 2.7.1 General ........................................................................................................ 58 2.7.2 Transfer of modelling results (A) ................................................................. 59 2.7.3 Creation of spatial data (B) .......................................................................... 60 2.7.4 Creation of flooding maps (C) ...................................................................... 62 2.7.5 Data sources ................................................................................................ 64


Guidelines Flood Risk Assessment ‐ ii ‐ May 2010

3 FLOOD DAMAGE ASSESSMENT ........................................................................................... 69 3.1 Introduction ................................................................................................................. 69 3.2 Classification of flood damages and benefits ............................................................... 69

3.2.1 Flood damages ............................................................................................ 69 3.2.2 Flood benefits .............................................................................................. 71

3.3 General approach to economic damage assessment ................................................... 72 3.3.1 Scope of the analysis and level of detail ..................................................... 72 3.3.2 Direct damage assessment .......................................................................... 73 3.3.3 Damage functions ........................................................................................ 74 3.3.4 Relative and absolute damage assessments ............................................... 76 3.3.5 Assessment of indirect damages ................................................................. 77

3.4 Damage models ............................................................................................................ 78 3.5 Economic and financial issues in damage assessment ................................................. 79

3.5.1 Financial vs. economic valuation ................................................................. 79 3.5.2 Estimating values of assets .......................................................................... 79 3.5.3 Time value of money ................................................................................... 80

3.6 Loss of life and health effects ....................................................................................... 80 3.6.1 Introduction ................................................................................................. 80 3.6.2 Loss of life .................................................................................................... 81 3.6.3 Health effects and social and environmental impacts ................................ 81

3.7 Presentation of damage assessments results .............................................................. 82 3.8 Data collection.............................................................................................................. 83 3.9 Flood damage assessment in the Mekong River Basin ................................................ 86

3.9.1 Scope and selected areas ............................................................................ 86 3.9.2 Approach and methodology ........................................................................ 86 3.9.3 Damage categories ...................................................................................... 87 3.9.4 Data collection and surveys ......................................................................... 87 3.9.5 Derivation of damage functions .................................................................. 92 3.9.6 Creation of damage maps ........................................................................... 95 3.9.7 Results of damage assessments .................................................................. 96

4 FLOOD RISK ASSESSMENT ................................................................................................ 103 4.1 Introduction ............................................................................................................... 103 4.2 General approach ....................................................................................................... 103 4.3 Presentation of the results of risk assessments ......................................................... 104

4.3.1 Economic risk ............................................................................................. 104 4.3.2 Risk to life .................................................................................................. 106

4.4 Risk reduction measures and risk evaluation ............................................................. 107 4.4.1 Effects of anti‐flooding and flood preparation measures on risk .............. 107 4.4.2 Basics of cost‐benefit analysis ................................................................... 108 4.4.3 Economic optimisation .............................................................................. 110 4.4.4 Risk valuation for loss of life ...................................................................... 111 4.4.5 Risk matrix ................................................................................................. 111

4.5 Risk assessment for the LMB ...................................................................................... 112 4.5.1 Introduction and approach........................................................................ 112 4.5.2 Results of risk assessment ......................................................................... 113

5 REFERENCES ..................................................................................................................... 117


Guidelines Flood Risk Assessment ‐ iii ‐ May 2010

APPENDICES Appendix 1 Approach to Flood Damage and Risk Assessment in Europe

Appendix 2 Approaches for the Assessment of Structural Damage to Buildings

Appendix 3 Fundamental issues in Economic Valuation

Appendix 4 Analysis of Human Instability in Flood Flows

Appendix 5 The Best Practice Guidelines and Project Phases/Stages

Appendix 6 Flow Chart to Create Flood Maps with a GIS

LIST OF FIGURES Figure 1‐1 General scheme for flood risk management. ................................................................. 5 Figure 1‐2 General scheme for flood risk assessment. .................................................................... 6 Figure 1‐3 Scheme for flood risk assessment used in the FMMP‐C2. ............................................. 6 Figure 2‐1 The sub‐basins or sub‐areas in the Lower Mekong River Basin. Letters refer to

the countries in which these sub‐basins are found (M= Myanmar, T = Thailand,

L = Lao PDR, V = Viet Nam, C = Cambodia). .................................................................. 12 Figure 2‐2 Overview of flood types in the Mekong River. ............................................................. 13 Figure 2‐3 EV1 and GEV‐fit to annual maximum daily rainfall at Pleiku for the years 1927‐

2006. ............................................................................................................................. 19 Figure 2‐4 Depth‐duration‐frequency curves for Pleiku. ............................................................... 20 Figure 2‐5 A 2‐hour‐design rainstorm diagram for the Upper Se San River derived from

measures used for Pleiku. ............................................................................................ 21 Figure 2‐6 Principles of the Clark Unit Hydrograph method, shown in various diagrams

that are related to each other. ..................................................................................... 22 Figure 2‐7 Hypothetical scatter plot of data points and fitted regression line showing the

relationship between flood volumes and flood peak water levels. The red

arrow in the Figure shows the deviation from the regression line for a single

observation. .................................................................................................................. 25 Figure 2‐8 TCEV‐distribution of the ratio of the T‐year return period event to the mean

annual flood for the pooled regional sample (Upper part of LMB, from

Adamson, 2007). .......................................................................................................... 27 Figure 2‐9 Fit of the GEV function to annual maximum daily discharge of the Mekong

River at Nakhon Phanom. ............................................................................................. 33 Figure 2‐10 Peak water level discharge‐Flood volume relation at Nakhon Phanom. ..................... 34 Figure 2‐11 Defined 2D grid of discharges (with return periods 2, 5, 10, 25, 50 and 100

years) and volumes (based on the regression line and lines of deviation from

the regression line). ...................................................................................................... 35 Figure 2‐12 Observed combinations of peak discharge and flow volume (red dots) and

selected 2D grid (blue dots). The green lines show the observed combinations

that most closely reflect the grid combinations. .......................................................... 35 Figure 2‐13 Example of scaling of an observed hydrograph to make it suitable for serving

as one of the grid points in Figure 2‐12. ....................................................................... 36 Figure 2‐14 Derived frequency curve for maximum annual water levels at a location in the

Mekong River floodplain just downstream of Nakhon Phanom. ................................. 37 Figure 2‐15 Development of the flood simulation matrix for combined floods. ............................. 41 Figure 2‐16 Hydrological map of Lower Xe Bang Fai River basin. .................................................... 42 Figure 2‐17 Impression of the 3‐D grid for peak discharge and flood volume in the Mekong

River (at Nakhon Phanom) and flood volume in the Xe Bang Fai River (at

Mahaxai). The blue dots only show the ‘outer side’ of the grid; the full grid

contains 90 such dots. .................................................................................................. 44 Figure 2‐18 Scaling of an observed hydrograph for the Xe Bang Fai River at Mahaxai. .................. 45


Guidelines Flood Risk Assessment ‐ iv ‐ May 2010

Figure 2‐19 GEV‐fit to marginal distributions of annual maximum discharge and flood

volume at Mahaxai (Xe Bang Fai River). ....................................................................... 46 Figure 2‐20 Flood volume – peak discharge relation for Nakhon Phanom and flood

volumes relation between Mahaxai and Nakhon Phanom, without and with Se‐

related lines. ................................................................................................................. 47 Figure 2‐21 Xe Bang Fai River hydraulic model with boundary conditions. .................................... 48 Figure 2‐22 Derived frequency curve for maximum annual water levels at a location in the

floodplain of the Xe Bang Fai River. ............................................................................. 49 Figure 2‐23 Flood depth and extent map of the Lower Xe Bang Fai River, for a return

period of T = 100 years. ................................................................................................ 50 Figure 2‐24 Hydraulic infrastructure of the Mekong Delta. ............................................................ 52 Figure 2‐25 The schematisation of the Mekong Delta using the hydrodynamic delta model

(ISIS) involves a large number of hydraulic nodes downstream of Kratie, next

to floodplain ‘reservoirs and river cross‐sections. ....................................................... 53 Figure 2‐26 Annual maximum water discharges at Stung Treng, 1910‐2006. ................................ 55 Figure 2‐27 Annual flood volumes at Stung Treng, period 1910‐2006. ........................................... 55 Figure 2‐28 Seasonal hydraulic inflow to the Tonle Sap Lake, estimated from 1910‐2006,

with recorded data for 1997‐2004. .............................................................................. 56 Figure 2‐29 Average monthly flow regime of the Mekong River at Stung Treng and of the

Tonle Sap Lake as percentage of the annual totals. ..................................................... 56 Figure 2‐30 Water level map, based on interpolation of ISIS nodes. .............................................. 58 Figure 2‐31 Water depth map, generated from a water level map subtracting the DTM

(ground) levels. ............................................................................................................. 58 Figure 2‐32 Depiction of how water or flood depth is obtained by subtracting the ground

level (DTM) from the water level (hydraulic simulation + GIS interpolation). ............. 59 Figure 2‐33 Example of ISIS data imported into Access, with node information and water

levels for different scenarios, various return periods (probabilities), and

different periods of the year. ....................................................................................... 60 Figure 2‐34 ISIS schematization. ISIS calculates with (among other data) reservoir nodes (+)

and cross‐section nodes (dots). A spatial component of ISIS contains cross‐

sections (blue lines) and reservoirs (purple polygons). ................................................ 61 Figure 2‐35 Example of a raster, with the reservoir raster (angular areas) superimposed on

the point‐based raster. ................................................................................................. 63 Figure 2‐36 Example of a flood depth map, here in the Mekong Delta for a flood with a

return period of 25 years. ............................................................................................ 64 Figure 3‐1 General approach to damage estimation, with the relationship between land

use and flood depths used to calculate damages. ....................................................... 73 Figure 3‐2 Example of a stage damage function for houses in the Netherlands (Kok et al.,

2005). Note that the curves for building damage and contents show these as

percentages of the total damage. ................................................................................ 75 Figure 3‐3 Schema of structural flood damage analysis. ............................................................... 76 Figure 3‐4 Absolute and relative damage approaches. ................................................................. 77 Figure 3‐5 Example of a mortality function derived for large‐scale flooding in the

Netherlands (Jonkman, 2007). ..................................................................................... 81 Figure 3‐6 Damage intensity per hectare of the flooding of South Holland province in the

Netherlands. ................................................................................................................. 83 Figure 3‐7 The absolute and relative flood damage assessment approaches require

different data sets. ....................................................................................................... 91 Figure 3‐8 District flood damage curves for three categories of damages in Koh Andet

District in Cambodia. .................................................................................................... 93 Figure 3‐9 Example of a damage function for housing, relating housing damage (as % of

the total value) to floodwater depths. ......................................................................... 94 Figure 3‐10 Relative flood damage curves for rice paddies, Cambodia. ......................................... 95


Guidelines Flood Risk Assessment ‐ v ‐ May 2010

Figure 3‐11: Steps (simplified) to create maps based on absolute and relative damage –

flood depth curves. ....................................................................................................... 97 Figure 3‐12 Example of a damage map based on relative damage curves, showing total

damage to rice paddy fields in 3 districts and their communes in Cambodia. ............. 99 Figure 3‐13 Example of a damage map based on relative damage curves, showing damages

to rice paddy fields per hectare in 3 districts in Cambodia and their communes. ..... 100 Figure 4‐1 Conceptual approach for determination of flood risk. ............................................... 104 Figure 4‐2 Relationship between return period and damage (left) and probability of

flooding and damage (right). ...................................................................................... 105 Figure 4‐3 The area below the curve, the total risk or expected damage (for example, in

USD/year), can be approximated by summing the areas of the rectangles, A, B,

b, C, c, D, d, E and e. ................................................................................................... 106 Figure 4‐4 Example of an FN curve. ............................................................................................. 107 Figure 4‐5 Effects of measures on flood risk. .............................................................................. 108 Figure 4‐6 Principle of economic optimisation. ........................................................................... 110 Figure 4‐7 Societal risk of fatalities due to flooding and the limit line of acceptable harm. ....... 111 Figure 4‐8 Structure of a risk matrix. ........................................................................................... 112 Figure 4‐9 Approach for estimation of the risk. .......................................................................... 112 Figure 4‐10 Probability damage curve for three Cambodian districts. .......................................... 113 Figure 4‐11 Flood risks by district for selected areas in Cambodia and Viet Nam. ....................... 114

LIST OF TABLES Table 2‐1 Overview of flood types per sub‐area (sub‐basin) ....................................................... 14 Table 2‐2 Parameters of EV1 and GEV distributions fitted to annual maximum daily

rainfall at Pleiku and rainfall values (mm) for selected return periods. ....................... 19 Table 2‐3 Short duration rainfall (mm) at Pleiku derived from daily rainfall extremes. .............. 20 Table 2‐4 Computation of design storm ordinates for 15 min intervals. ..................................... 21 Table 2‐5 Parameters of the GEV distribution fitted to annual maximum daily discharge

of the Mekong River at Nakhon Phanom and discharges (m3/s) for selected

return periods. .............................................................................................................. 34 Table 2‐6 Average monthly and annual flow (MCM) of the Xe Bang Fai River at Mahaxai

and Mekong River at Nakhon Phanom. ........................................................................ 42 Table 2‐7 GEV‐parameters, flood peak discharges and flood volumes (June‐November)

for distinct return periods in the Xe Bang Fai River at Mahaxai. .................................. 44 Table 2‐8 Overview of water level monitoring stations at coastal boundaries in Viet Nam. ....... 57 Table 3‐1 Classification of flood damages. ................................................................................... 69 Table 3‐2 Distribution of damage over damage categories for a flood scenario for South

Holland province in the Netherlands (Jonkman et al., 2008). ...................................... 83 Table 3‐3 Overview of Flood Damage Data .................................................................................. 84 Table 3‐4 Observed rank share of damages in the focal areas .................................................... 87 Table 3‐5 Data sources used for mapping damages in the Cambodian Mekong Delta. .............. 90 Table 3‐6 Damage data for different return periods for Koh Andet District. Damage is

shown in million USD. ................................................................................................... 96 Table 4‐1 Illustrative summary of damage and probability calculations for a range of

flood scenarios. .......................................................................................................... 104 Table 4‐2 Results of risk calculations for the three focal areas in Cambodia. ............................ 113

MRC Flood

Guidelines

d Management a

Flood Risk Asse

and Mitigation P

essment

Programme Commponent 2: Struc

‐ vi ‐

ctural Measures and Flood Prooffing

May 2010


Guidelines Flood Risk Assessment ‐ vii ‐ May 2010

ABBREVIATIONS AND ACRONYMS ADCP Acoustic Doppler Current Profiler (Acoustic Doppler Profiler); an

instrument used to measure the velocity of water across an entire water column

ArcGIS ArcGIS is a complete software system for designing and managing solutions through the application of geographic knowledge

ARF Area Reduction Factor (hydrology) BCM Billion Cubic Meters BDP Basin Development Planning BPG Best Practice Guidelines CBA Cost Benefit Analysis d/s Downstream DACA Damage and Casualties Assessment project for the Lower Mekong

Basin using the HIS‐SSM for assessing damages and casualties DEM Digital Elevation Model (see also DTM), is a digital model or 3D

representation of a terrain's surface created from terrain elevation data

DSF Decision Support Framework DTM Digital Terrain Model (see also DEM) EC European Commission EU European Union EV1 Extreme Value type 1 distribution (hydrology) EXCIMAP European Exchange Circle on Flood Mapping FEMA Federal Emergency Management Agency of the USA FHA Flood Hazard Assessment FMM Flood Management and Mitigation FMMP‐C2 Flood Management and Mitigation Programme, Component 2

(MRC) FN curve Societal risk is usually defined as the relationship between the

probability of a catastrophic incident, expressed as the average frequency with which it can be expected to occur, and its consequences. It is usually represented as an F‐N curve in which F is the annual frequency and N represents the number of casualties.

FRA Flood Risk Assessment FV Future Value (economic analysis) GEV Generalised Extreme Value distribution (hydrology) GIS Geographic Information System HAZUS Software model used for risk assessment analysis of potential losses

from floods, hurricane winds and earthquakes (by FEMA) HH Household(s) HIS‐SSM Hydrological Information System: a damages and casualties

assessment module HYMOS An information system for water resources management IDW Inverse Distance Weighting method: an interpolation method to

obtain a continuous GIS‐raster on the basis of data points (nodes), assigning the most weight to nearby points by using their distance to the point to calculate(see also NN)

IFRM Integrated Flood Risk Management ISIS Hydrodynamic simulator for modelling flows and levels of water in

open channels and estuaries IUH Instantaneous Unit Hydrograph (hydrology)


Guidelines Flood Risk Assessment ‐ viii ‐ May 2010

JICA Japan International Cooperation Agency LMB Lower Mekong (River) Basin LMD Lower Mekong Delta LXQ Long Xuyen Quadrangle (Viet Nam) MCM Million Cubic Meters MRC(S) Mekong River Commission (Secretariat) MRB Mekong River Basin MSL Mean sea level, the average (mean) height of the sea, with

reference to a suitable reference surface NN Natural Neighbours method: an interpolation method to obtain a

continuous GIS‐raster on the basis of data points (nodes), assigning the greatest weight to nearby points by calculating overlapping areas in Voronoi/Thiessen polygons (see also IDW)

NPV Net Present Value (an economic term) PDR (Lao) (Lao) People’s Democratic Republic PoR Plain of Reeds (Viet Nam) PV Present Value (an economic term) RFMMP Regional Flood Management and Mitigation Programme RID Royal Department of Irrigation RR Rainfall Ratio (hydrology) SBF Xe Bang Fai River (Lao PDR) SCS‐CN Soil Conservation Service (USA) Curve Number method (hydrology) SWAT Hydrological model: river basin scale model quantifying the impact

of land management practices in large, complex watersheds TCEV Two Component Extreme Value (hydrology) u/s upstream UH Unit Hydrograph (hydrology) UK United Kingdom UNESCO‐IHE Institute for Water Education (IHE) of the United Nations

Educational, Scientific and Cultural Organization (UNESCO) USA United States of America UTM The Universal Transverse Mercator is the geographic coordinate

system, which uses a 2‐dimensional Cartesian coordinate system to give locations on the surface of the Earth

VRSAP The “Vietnam River Systems and Plains” model is a hydro‐dynamic model to simulate mainly one‐dimensional hydrodynamic river flow or quasi two‐dimensional flow on floodplains

WUP Water Utilisation Programme WUP‐A Water Utilisation Project Component A: Development of Basin

Modelling Package and Knowledge Base


Guidelines Flood Risk Assessment ‐ ix ‐ May 2010

GLOSSARY Attribute data Also known as "count" data.

Areal rainfall Rainfall in a specific area expressed as an average depth of liquid

water over the area; the average rainfall over the specific area calculated on timescales on a storm, seasonal, or annual basis.

Backwater The rise in surface elevation of flowing water upstream from and as a result of an obstruction to flow. The difference between the observed stage and that indicated by the stage‐discharge relation is reported as backwater.

Best Practise Guidelines Best Practice Guidelines are guidelines, describing methods or techniques that have consistently shown results superior to those achieved with other means, and that are used as a benchmark. Best Practice Guidelines are also interpreted guidelines describing processes of developing and following a standard way of doing things that dedicated organizations can use.

Colmatage The natural deposition of particles suspended in water on land areas, usually on river floodplains and deltas, coastal lowlands, and flooded meadows.

Convective rainfall Occurs when rain particles form in the active updraft of a cumulonimbus cloud, grow primarily by the collection of cloud droplets (i.e., by coalescence and/or riming) and fall out not far from their originating updraft.

Damage curve The functional relationship between inundation characteristics (depth, duration, flow velocity) and damage for a certain category of risk.

Design rainstorm A critical rainfall event that is used for assessing the flood hydrograph of a certain return period.

Direct damage All harm which relates to the immediate physical contact of floodwater to people, property and the environment. This includes, for example, damage to buildings, productive assets, loss of crops and livestock, loss of human life, immediate and immediately foreseeable and calculable health impacts to the population in the flooded area and ecological harm.

Double mass test A statistical test that is used to check the consistency of many kinds of hydrologic data by comparing date for a single station with that of a pattern composed of the data from several other stations in the area.

Exposure A measure of the people, assets and activities threatened by a flood hazard.

Flashiness Term reflects the frequency and rapidity of short term changes in streamflow, especially during runoff events.


Guidelines Flood Risk Assessment ‐ x ‐ May 2010

Flood control A structural intervention to reduce a flood hazard.

Flood damage Damage to people, property and the environment caused by a

flood. It includes direct as well as indirect damages.

Flood damage map Map of the predicted flood damage.

Flood damage risk (= Flood risk)

The combination or mathematical product of the probability of the flood hazard and the possible damage that it may cause. This risk can also be expressed as the average annual possible damage or expected damage. If it is expressed in a financial measure, it is the expected net present value using economic valuations with different measurement assumptions.

Flood hazard A flood that may potentially result in damage. A hazard does not necessarily lead to damage.

Flood hazard map Map of the predicted or documented extent/depth/velocity of flooding with an indication of the flood probability.

Flood proofing A process for preventing or reducing flood damages to infrastructure, buildings and/or the contents of buildings located in flood hazard areas.

Flood risk management Comprehensive activity involving risk analysis, and both identification and implementation of risk mitigation measures.

Flood risk management measures

Actions that are taken to reduce the probability of flooding or the possible damages due to flooding or both.

Flood risk map Map of the predicted extent of different levels/classes of average annual possible damage due to flooding.

Hydrograph A time series of water levels at a fixed location. For floods, it shows the rise and fall of flood waters and the peak water level height of the flood.

Hydrological hazard A hydrological event (discharge) that may result in flooding.

Hydraulic roughness Is the measure of the amount of frictional resistance water experiences when passing over land and channel features.

Hydrodynamic load The pressures that result from water flowing against and around a rigid structural element or system. The pressures are functions of velocity, direction of flow relative to the object, object geometry, and object surface roughness characteristics.

Hydrostatic load The pressure at any water depth due to hydrostatic pressures.

Hyetograph A graphical representation of the distribution of rainfall over time.

Indian 1954 A geodetic datum first defined in 1954 and used by MRC for the


Guidelines Flood Risk Assessment ‐ xi ‐ May 2010

Mekong River Basin.

Indirect damage All non‐direct damage which relates to the disruption of economic activity and services due to flooding including impacts on the non‐flooded areas.

Integrated flood risk management

The approach to flood risk management that focuses on the full chain of a meteorological hazard leading to flood damages and considers combinations of structural and non‐structural solutions to reduce that damage.

Meteorological hazard A meteorological event (storm) that may result in a hydrological hazard and, eventually, in flooding.

Monte Carlo analysis A computer sampling technique often used for simulating long term events based on statistical probabilities, testing a series of multiple random events by drawing random samples from a data set.

Non‐staggered grid A grid for which vector variables and scalar variables are stored at the same locations.

Orographic effect Occurs when an air mass approaches a mountain range and is rapidly forced upward, causing any moisture to cool and create precipitation in the form of rain or snow.

Unit hydrograph (UH) The hypothetical unit response of a watershed (in terms of runoff volume and timing) to a unit input of rainfall. It can be defined as the direct runoff hydrograph resulting from one unit (e.g., one cm or one inch) of effective (net) rainfall occurring uniformly over that watershed at a uniform rate over a unit period of time.

Raster Defines values for pixels occupying a particular rectangular area of the plane, not necessarily including (0, 0).

Rating curve A rating table or curve is a relationship between stage (water level) and discharge at a cross section of a river.

Reach Any length of a stream between any two points.

Resilience The ability of a human social system (at the level of community or society) to cope with the damaging effect of floods.

Return period Recurrence time, average time interval between subsequent events in which conditions are exceeded. When designing a structure, the return period is usually larger than the projected lifetime, because, for instance, if both would equal 50 years, the structure would have a 64% probability of failure during its lifetime. In statistical analysis an event with a return period of N years is likely, on average, to be exceeded only once every N years.

Routing A technique used to predict the changes in shape of water as it moves through a river channel or a reservoir.


Guidelines Flood Risk Assessment ‐ xii ‐ May 2010

Saint Venant equations The shallow water equations in its unidimensional form are a set of hyperbolic partial differential equations that describe the flow below a pressure surface in a fluid (sometimes, but not necessarily, a free surface).

Scour Washing away of the bed/bank material under the action of current and wave.

Serial and cross correlation

The existence of serial correlation in a time series will affect the ability of the test to assess the site significance of a trend; and the presence of cross‐correlation among sites in a network will influence the ability of the test to evaluate the field significance of trends over the network.

Spatial data Data or information that identifies the geographic location of features and boundaries on Earth, such as natural or constructed features, oceans, and more. Spatial data is usually stored as coordinates and topology, and is data that can be mapped. Spatial data is often accessed, manipulated or analyzed through Geographic Information Systems.

Spot height The height of the highest point in a given area expressed in feet or meters above sea level, as marked on topographical charts. These are indicated by black dots with adjacent numerals.

Susceptibility The opposite of resilience: the inability of a human social system (at the level of community or society) to cope with the damaging effect of floods.

(N‐year) synthetic series Simulated long term event‐related data series.

Vulnerability The potential damage that flooding may cause to people, property and the environment.

MRC Flood Ma

Guidelines Flo

REFEREN The Flood Msymbols in t A. Type oB. Projec The goal ofapplications A) Text/C

i) P

Fo

ii) T

T

iii) EE

iv) A

MRin

B) Projec

coverssubdiv

a) P

b) F c) P

d) D e) C

Any part of of the five s Sometimes

anagement and

ood Risk Assessm

NCE SYMB

Managementthe left marg

of text/content stages with

f the symbos or to try to

Content Symb

roject backgrMMP‐projectr explanation

heory he theory be

xample xamples of th

pplications aMethodology River Basin (LMn one of the f

t Stage Symbs only the vided into five

reliminary/p

easibility stud

reliminary de

Detailed desig

Construction/

a guideline fstages is mar

more than o

Mitigation Prog

ment

BOLS FOUN

and Mitigatigins for quick

nt; and in the five co

ls is to help find informat

bols: The repo

round/Reportt informationn of the repor

hind the pro

he proposed/

nd Guidelineand theory aMB), includinfive project st

bols: A projecsecond phae stages:

re‐feasibility

dy and overa

esign stage

gn and detaile

bid documen

falling outsideked with a w

one symbol m

gramme Compon

ND IN THE

ion Programmreference. T

onsecutive pr

readers to stion needed

ort texts are

rt informationn and backgrort structure o

posed/applie

/applied met

s adapted/applng guidelines.tages describ

ct usually conase: Plannin

study stage

all planning st

ed planning s

nts stage

e the scopehite cross:

may apply to a

nent 2: Structur

‐ xiii ‐

E TEXT

me Componehe symbols a

roject phases

scan the textfor work at s

categorised i

n ound, or content.

ed methods a

hods and gui

ied to the Lo The guidelinbed below (B)

nsists of five png/Developm

tage

stage

a section.

al Measures and

nt 2 (FMMP‐re of two typ

.

t to look for pecific projec

nto four grou

nd guidelines

delines.

wer Mekong es are applie.

phases (see Sent/Design.

d Flood Proofing

‐C2) guidelinepes. They ind

theory, exact stages.

ups, as follow

s.

g ed

Section 1.6). FThis phase

g

May 2010

es contain icate:

mples, or

ws:

FMMP‐C2 can be

CHA

APTEER 11

INTROODUCTTION


Guidelines Flood Risk Assessment ‐ 3 ‐ May 2010

1 INTRODUCTION

1.1 Guide to the reporting structure of the Flood Management and Mitigation Programme ‐ Component 2, Structural Measures and Flood Proofing

The Mekong River Commission (MRC) implemented Component 2 on Structural Measures and Flood Proofing of the Flood Management and Mitigation Programme (FMMP) between September 2007 and January 2010 under a consultancy services contract between the Mekong River Commission Secretariat (MRCS) and Royal Haskoning, working in association with Deltares and UNESCO‐IHE. The work comprised three periods, an Inception period and two Implementation periods. During each period, the consultants delivered a series of outputs and discussed them with the MRC, the National Mekong Committees, and line agencies of the four MRC Member Countries. A portion of Component 2 ‐ on 'Roads and Floods' ‐ was implemented by the Delft Cluster under a separate contract with the MRC. Component 2 included five Demonstration Projects which are presented separately. The consultancy services contract for Component 2 requests, in general terms, four main products in addition to a Final Report. The reports produced as of the completion of Component 2 (FMMP‐C2) are structured as follows: Volume 1 Final Report

Volume 2 Characteristics of Flooding in the Lower Mekong Basin Volume 2A Hydrological and Flood Hazards in the Lower Mekong Basin Volume 2B Hydrological and Flood Hazards in Focal Areas Volume 2C Flood Damages, Benefits and Flood Risk in Focal Areas Volume 2D Strategic Directions for Integrated Flood Risk Management in Focal Areas

Volume 3 Best Practice Guidelines for Integrated Flood Risk Management Volume 3A Best Practice Guidelines for Flood Risk Assessment Volume 3B Best Practice Guidelines for Integrated Flood Risk Management Planning and

Impact Evaluation Volume 3C Best Practice Guidelines for Structural Measures and Flood Proofing Volume 3D Best Practice Guidelines for Integrated Flood Risk Management in Basin

Development Planning Volume 3E Best Practice Guidelines for the Integrated Planning and Design of Economically

Sound and Environmentally Friendly Roads in the Mekong Floodplains of Cambodia and Viet Nam1

Volume 4 Project development and Implementation Plan Volume 5 Capacity Building and Training Plan Demonstration Projects Volume 6A Flood Risk Assessment in the Nam Mae Kok Basin, Thailand Volume 6B Integrated Flood Risk Management Plan for the Lower Xe Bang Fai Basin, Lao

PDR Volume 6C Integrated Flood Risk Management Plan for the West Bassac Area, Cambodia Volume 6D Flood Protection Criteria for the Mekong Delta, Viet Nam Volume 6E Flood Risk Management in the Border Zone between Cambodia and Viet Nam This report is Volume 3A in the above series.

1 Developed by the Delft Cluster

MRC Flood

Guidelines

The FMMsteps: thPhase beReferencmonths abe carriePhase. S

1.2

The MRCManagemthe MRC

Polic

Strat

Proje for flood Each Meguide or complem The BPGintendedthe LMBcountry a The basisThe Bestguideline

1.3

1.3.1

In generamanagemconseque ‘"Flood radverse activity a In this codanger, iin case o (Integratmanage managem

d Management a

Flood Risk Asse

MP Componehe Inception egan at the ece with a Regafter project ed out in a See for refere

Best Practi

C Secretariatment under C and Membe

cy formulatio

tegy and plan

ect design an

risk manage

ember Countrregulate the

ment the BPG

Gs do not atd as a recipe . Rather the and project c

s of flood rist Practice Guies.

Concepts o

Flood risk a

al, the word ment it denences. The de

risk" means consequenceassociated wi

ontext, the tei.e. the (probof an event.

ted) Flood risflood risks inment consisti

and Mitigation P

essment

ent 2, StructPhase and Send of Septegional Worksinitiation. Thperiod of 6 ence Final Rep

ice Guideline

t has developFMMP‐C2 toer Country lin

n;

nning for poli

nd evaluation

ement in the L

ry of the MRe planning, evs.

tempt to sufor carrying BPGs provid

context.

sk managemeidelines for fl

of Flood Risk

nd flood risk

‘risk’ refers totes the coefinition ado

the combinaes for humaith a flood eve

erms hazard aability of) flo

sk managemen a given riveing of four co

Programme Com

tural MeasurStages 1 andember 2007 shop in Ho Ce original TOmonths, leavport, Volume

es for Integr

ped Best Prao provide pole agencies w

cy implemen

;

Lower Mekon

C already havaluation and

mmarise or out planning

de an inform

ent in the LMood risk asse

k Assessmen

managemen

to the probaombination opted by the E

tion of the pn health, thent.’

and vulneraboding. Vulne

ent is an apper system. Figomponents.

mponent 2: Struc

‐ 4 ‐

res and Flood 2 of the Imand concludChi Minh CityOR envisaged ving 12 mone 1.

ated Flood R

actice Guidellicy‐makers,

with a commo

ntation;

ng Basin (LM

as its own pod implement

replace thesg or project ation resourc

MB is proper essment (BPG

t in the Best

nt

bility of loss of the probaEuropean Com

probability ofhe environme

ility are oftenerability relate

proach to idegure 1‐1 pres

ctural Measures

d Proofing, mplementatioed in accordy at the end the Stage 1 I

nths for the

Risk Manage

ines (BPG) fomanagers ann standard to

B).

licy and set otation of floo

se national gdesign for floce/tool that

assessment oG for FRA) in

t Practice Gu

or harm. In ability of a mmission is t

f a flood eveent, cultural

n used. Hazares to potentia

entify, analyssents a gene

and Flood Proof

was developon Phase. Thdance with thof January 2mplementatiStage 2 Imp

ement

or Integratednd FMM profo apply in:

of legal framod risk manag

guidelines, nood risk mancan be adap

of the managthe LMB offe

uidelines

the context oflood with he following:

ent and of thheritage an

rd refers to thal adverse co

se, evaluate, ral scheme fo

fing

May 2010

ped in three he Inception he Terms of 2008, only 4 ion Phase to lementation

d Flood Risk fessionals in

eworks that gement that

nor are they nagement in pted to each

geable risks. er one set of

of flood risk its adverse :

he potential nd economic

he source of onsequences

control and or flood risk

MRC Flood Ma

Guidelines Flo

Definiti

Quantitmeasur

Evaluat

Identifistructuinspect

Quafl

R

S

Figure 1‐1

This generacould also minimisatiothe approamaximise th The concepthe quantitstep. Otherevaluation s 1.3.2 F

The identifseveral stepcomponent 1) Flood S

terrain 2) Flood

meteorhydrolovolume

3) Vulneraeconom

anagement and

ood Risk Assessm

ion of the flo

tative analysre, presented

tion (assessm

cation of mral measuretion or mainte

ntitative anaood risk ma

Risk Evalua

Risk reductiocontrol

System defiScope, haz

General sch

al scheme focbe used to

on of droughach can focuhe performan

pt of flood risative analysisr guidelines/step (see Sec

lood risk asse

fication and ps. Figure 1‐2ts:

System definand water flohazard analyrological eveogical hazardes). The resultability and dmic data and

Mitigation Prog

ment

od system, th

sis of probad graphically o

ment) of risks

easures for es as well aenance).

alysis and pping

ation

on and

nition zards

heme for flood

cuses on minio assess thet, maximisatus on multipnce of the wa

k assessments of the level /reports offection 1.2 for a

essment

mapping of f2 presents a g

nition and coows). ysis of the onts that mayd characterists can be evedamage assea vulnerabilit

gramme Compon

he hazards an

bilities and or displayed

acceptable o

risk reductioas managem

Structunon-str

meas

risk manageme

imisation of fe overall hytion of waterle objectivesater system.

t generally re of flood risk er insight intan overview o

flood risks regeneral schem

ollection of b

occurrence oy eventually stics of the entually displaessment in tty/damage m

nent 2: Structur

‐ 5 ‐

nd the scale a

adverse conon a flood ris

or unacceptab

on and contrment and co

-

ral and ructural sures

ent.

flood risks to ydrological pr quality and s: not only t

efers to the sin an area oro possible mof other guide

equires seveme for flood

basic data (e

of flooding. lead to floorespective wayed by meathe areas pr

model.

al Measures and

and scope of

nsequences csk map.

ble; and

rol, includingontrol optio

an acceptabperformance ecological qto minimise

econd step inr basin. This rmeasures andelines).

ral types infrisk assessm

.g., basic dat

This step incoding combinwatershed (pens of flood (hrone to flood

d Flood Proofing

the analysis.

combined in

g structural ns (e.g., mo

ble level. The of the syst

quality). In suthe risk, bu

n this generareport focused approache

formation coent, consistin

ta on elevati

cludes an anned with dateak flood dihazard) mapsding based

g

May 2010

to a risk

and non‐onitoring,

approach tem (e.g. uch cases, ut also to

l scheme; es on that es for the

llected in ng of four

on of the

nalysis of ta on the ischarges, s. on socio‐



4) Risk determination and flood risk mapping combining the results of the flood hazard analysis, (the probability of a certain hazard) with the results of the damage assessment. These results can be displayed in risk maps, graphically, or numerically in ways that offer insight into the expected annual damage.

These steps are presented in separate sections in this report.

Figure 1‐2 General scheme for flood risk assessment.

The FMMP‐C2 project uses a variation on this scheme that is shown in Figure 1‐3. The figure indicates how meteorological information is combined with watershed and river information to determine hydrological and flood hazards. Damages can be determined by combining data for the possibility of floods with data on the vulnerability to damage of the affected area. When the return period data for different flood events are included, the probability of a certain type of damage ‐‐ i.e., the risk – can be determined. The figure also indicates how crisis management can reduce the risk at the final stage.

crisis management

precipitation peak discharges volumes

floods damage

meteorological hazard

hydrological hazard

watershed

river channel

flood hazard

damage risk

damagevulnerability

hazard risk

crisis management


floods damage


hydrological hazard

watershed

river channel

flood hazard

damage risk

damagevulnerability

hazard risk


floods damage


hydrological hazard

watershed

river channel

flood hazard

damage risk

damagevulnerability

hazard risk

Figure 1‐3 Scheme for flood risk assessment used in the FMMP‐C2.

Flood hazard maps

System definition and collection of basic data

Flood damage maps

Risk determination and flood risk mapping

Vulnerability and damage assessment

Flood hazard analysis

Flood risk maps



1.4 Basis of the Flood Risk Assessment Guidelines

The Best Practice Guidelines in the area of Flood Risk Assessment on the basis of: 1. a review of other guidelines or ‘best practice’ documents in the field of flood risk

assessments (see Appendix 1 of this BPG); 2. a review of existing guidelines in the LMB countries regarding damage assessments; and 3. experience in flood risk assessments in the LMB focal areas during Stage 1 of the FMMP‐C2. During Stage 1 of the FMMP‐C2, a flood damage data collection and processing methodology was designed for the LMB focal areas that takes into account the availability of secondary data as well as time and budget constraints. Survey teams collecting secondary and primary data offered feedback on the feasibility of the methodology along with consultants. Both concluded that the methodology was feasible and suitable for use in the MRC Member Countries and that the damage data collected could be used for estimating flood risks. (For details on the FMMP‐C2 damage data collection, see Annex 2 to the Stage 1 Evaluation Report.) The details and results of these methodologies applied for the assessment of the flood hazards in the LMB focal areas were presented in Annex 1 of the Stage 1 Evaluation Report and were updated in Stage 2. The guidelines for flood risk assessment have been tested and amended during Stage 2 of FMMP‐C2, especially through the demonstration projects. The methodology was also discussed in regional training courses.

1.5 Use of the Flood Risk Assessment Guidelines

Who should use the “Best Practice” guidelines and for what purpose? The four target groups for these guidelines are: 1. Policy makers who will participate in the development of the flood risk management

strategies at national and/or regional levels. 2. Regional and/or basin planners who should be aware of the impact of development

scenarios on flood risks and who should apply flood risk management as an essential component of Integrated Water Resources Management. The IFRM BPG for Basin Development Planning (BDP) is also relevant to this category of planners.

3. Flood risk management related project designers, for structural or non‐structural projects.

They can use these guidelines to estimate field level flood benefits and costs to input in project analysis as described in the Best Practice Guidelines for IFRM Planning and Impact Assessment.

4. Participants in multilateral discussions on transboundary flood risk related impacts. The BPG for FRA provides a set of methodologies for assessing: 1. flood hazards; 2. flood damages; and 3. flood risks.

MRC Flood

Guidelines

These guThis docdescribesand vulnEach sect

Prin

Sele

Pro

Exa These guassessmeand modsteps tha Precise escope ofexample,floods inprocessehazard a

1.6

In order projects 1. Initi2. Plan3. Prod4. Mon5. Clos The BesDevelopmare elabo

d Management a

Flood Risk Asse

uidelines comcument follos the processerability assetion is compr

nciples, conce

ected models

posed approa

mples of resu

uidelines canent in the LMdels in risk assat need to be

elaboration af the assessm, there are sen the floodplaes such as flonalysis.

The Best P

to facilitate into phases s

iation nning/Develoduction/Implnitoring/Contsure

st Practice ment/Designorated in App

and Mitigation P

essment

mbine theoryws the genes of analysingessment. Sectrised of four s

epts and term

s (for defining

aches for the

ults from Stag

n be adoptedMB. These gusessment, bue followed for

nd applicatioment, availabeveral types oains (see Secood duration

ractice Guid

managementsuch as the fo

opment/Desiglementation trol

Guidelines a. This phase,pendix 5 and

Programme Com

and practiceeral scheme g flood hazartion 4 showssub‐sections.

minology;

g and quantify

e LMB; and

ge 1 of the M

d as part of aidelines do nut focus, inster a complete

on of risk asseble data and of flooding inction 2.2). Eaor flow spe

elines and p

t of an enginollowing five

gn

are almost its stages anare presente

mponent 2: Struc

‐ 8 ‐

e with speciffor flood ri

rds. Section 3s how flood r. These prese

fying hazards,

Mekong FMM

a consistent not fully presead on the prrisk assessme

essment dep local conditn the LMB: triach type of fed. These, a

project phase

neering projephases:

exclusively nd the assoced briefly, abo

ctural Measures

fic emphasis isk mapping

3 presents theisk can be deent:

, consequenc

P‐C2.

and uniformsent all of therinciples and ent.

ends on mantions (e.g., hbutary floodsflooding is asmong other

es/stages

ect, project m

applicable ciated symboove, in the re

and Flood Proof

on flooding (Figure 1‐2)e tools for floetermined an

ces or risks);

m approach te details of toffer an ove

ny factors: obydrology, lans, main streassociated witfactors, affe

managers nor

to Phase 2ls used in thport introduc

fing

May 2010

in the LMB. ). Section 2 ood damage nd displayed.

to flood risk the methods erview of the

bjectives and nd use). For m floods, or th particular ct the flood

rmally divide

2: Planning/ e guidelines ction.

CHA

APTEER 2FLO

2

OOD HAZARD ASSSESSMENT

MRC Flood Ma

Guidelines Flo

2 F

2.1 In

There are hazards. Wavailable mmakers muquestions to

What t

From w

What q

Which

What This type ofguidelines iclassificatiobecause thguidelines

flooding tha

in the MRB Section 2.7

2.2 C

A classificatincluding:

Tributa

Main s

Combi

Floods

Floods The LMB isoverview of The flood ty Tributary flfloods, causquickly. Flaeffects of fFurther dowlevels are n MainstreamThe hydrolmaximum w

anagement and

ood Risk Assessm

LOOD HAZA

ntroduction

a variety of Whether or nmeasures) in st have a cleo start with a

types of flood

where does t

quantities of

factors influ

is the freque

f analysis is ren flood hazaon list for the he different that are app

at are listed i

.

describes flo

Classification

tion of floods

ary floods,

stream floods

ined floods,

s in Cambodia

s in the Meko

divided in sf the location

ypes are desc

loods occur ised by intensash floods arflash floods, wnstream, inot affected b

m floods are ogical hazarwater levels.

Mitigation Prog

ment

ARD ASSESS

measures aot these meplace. In ordar understanare:

ds occur?

he water com

water are inv

ence water f

ncy of floodin

eferred to asrd assessmendifferent typtypes requi

plicable to es

n Section 2.2

ood hazard m

n of type of f

s is used as a

s,

an Floodplain

ong Delta.

ub‐basins nun where diffe

cribed in deta

n the steep sse rainfall aftere short livedwhen accomn the middleby backwater

floods alongd is determFlood volum

gramme Compon

SMENT

available to easures are eder to choosnding of the p

me and to wh

volved?

flows?

ng?

s “flood hazarnt for the Mepes of flood eire differentstimating flo

2 along with

mapping, follo

flood

applied by the

n, and

umbered fromrent flood typ

ail as follows.

sloped upperer a long raind, rise and fmpanied by lae sections offrom the Me

g the Mekongined by the es are impor

nent 2: Structur

‐ 11 ‐

prevent or reffective stroe the appropprocesses of f

here does it fl

rd assessmenekong River Bevents in the t guidelines. ood hazard f

examples of

wing the asse

e MRC in the

m 1 to 10 (sepes occur in t

.

r reaches of tny period andfall rapidly aandslides, aref tributaries ekong, as in t

g caused by flood peak

rtant for mea

al Measures and

reduce damaongly dependpriate protecflooding invo

low?

nt”. This chapBasin (MRB). SMRB. This claSections 2.or each of t

their use for

essment proc

Lower Meko

ee Figure 2‐1the LMB.

he basins. Thd forcing the cnd have highe equivalent the flashineshe Nam Mae

high water lewater levelssuring flood

d Flood Proofing

aging effectsds on the syctive measurolved. Some o

pter presentsSection 2.2 passification is.3 ‐ 2.6 prethe different

selected “foc

cedures.

ong River Bas

). Figure 2‐2

hey are genecatchment toh flow velocto dam breass reduces, e Kok at Chian

evels on thes responsibleduration. Ma

g

May 2010

of flood ystem (of es, policy of the key

the basic presents a s relevant esent the types of

cal areas”

sin (LMB),

offers an

rally flash o respond ities. The ak waves. but flood ng Rai.

Mekong. e for the ainstream



floods occur at a number of locations along the Lower Mekong River all the way from Chiang Saen to Kratie.

Figure 2‐1 The sub‐basins or sub‐areas in the Lower Mekong River Basin. Letters refer to the countries in

which these sub‐basins are found (M= Myanmar, T = Thailand, L = Lao PDR, V = Viet Nam, C =

Cambodia).

A Northern highlands 1 Northern Laos 2 Chiang Rai B Central plateau and highlands 3 Nong Khai / Songkhram 4 Central Laos 5 Mun / Chi C Southeast highlands 6 Soutthern Laos 7 SeSan / Sre Pok / Se Kong D Southern region 8 Kratie 9 Tonle Sap 10 Delta

1L

1V

2T

3L

3T

4L

5T 6L

6C

7L

7V

7C

8V

9C9T

10C

10V

Areas outside the Lower Mekong Basin but affected by Mekong floods

8C

MRC Flood Ma

Guidelines Flo

Figure 2‐2

Combined fdeterminedup the tribubackwater may continseveral low Floods in thPenh and inand out fro

anagement and

ood Risk Assessm

Overview o

floods are flod by a combiutary levels amay flow intue for weeks

wer reaches of

he Cambodianclude floodm the Tonle

Mitigation Prog

ment

of flood types in

oods in the dnation of triband impede do the tributas. In the shallof tributaries f

an floodplaining around tSap Lake.

gramme Compon

n the Mekong R

downstream butary flow adrainage. Wharies. These fow areas aloface this type

n occur alonghe Tonle Sap

nent 2: Structur

‐ 13 ‐

River.

sections of tand water levhen water levfloods are nong the Mekoe of flooding.

g the Mekongp Lake and th

al Measures and

ributaries, wvels in the Mvels in the Mt short term ng River dow

g downstreahe Tonle Sap

d Flood Proofing

here the flooekong River

Mekong River like flash flo

wnstream of V

m of Kratie tp River that f

g

May 2010

od level is that back are high,

oods; they Vientiane,

to Phnom flows into



Floods in the Mekong Delta occur in the Mekong and Bassac Rivers and their floodplains, including the areas of colmatage canals that divert and control the flow from and to the rivers. In the Mekong Delta water levels rise slowly due to storage in the Tonle Sap Lake and in the Mekong River floodplains. Flooding here is recognised as essential for soil fertility, biodiversity and aquaculture. At the same time, it hampers use of agricultural land. The flood levels in the Mekong Delta downstream are essentially the result of upstream and lateral inflow and downstream water levels of the sea. Table 2‐1 offers an overview of flood hazards of these different types of floods in the sub‐areas of the LMB.

Table 2‐1 Overview of flood types per sub‐area (sub‐basin)

Sub‐area Type of flood Issue Assessment of hydrological

hazard Assessment of flood hazard

SA1 Northern Laos

Tributary Mainstream Combined

yes yes no

From rainfall statistics, discharge statistics or regional approach Annual Flood Report ‐

Rainfall‐runoff + hydraulic model/satellite imagery Hydraulic model/satellite imagery ‐

SA2 Northern Thailand


yes yes yes

From rainfall statistics, discharge statistics or regional approach Annual Flood Report HYMOS/other databases

Rainfall‐runoff + hydraulic model/satellite imagery Hydraulic model/satellite imagery As above

SA3 Nong Khai/ Songkhram


yes yes yes

For some basins via statistics rest via regional approach Annual Flood Report HYMOS/other databases

Models available for Nam Loei, Huai Mong and Nam Songkhram/ satellite imagery Hydraulic model/satellite imagery As above

SA4 Central Laos


yes yes yes


Models available for Nam Ngum, Xe Bang Fai and Se Bang Hieng/ satellite imagery Hydraulic model/satellite imagery As above

SA5 Mun‐Chi


yes yes no

For 12 via statistics, rest via RID data or regional approach HYMOS/RID/other databases Annual Flood Report ‐

Hydraulic models/satellite imagery Hydraulic model/satellite imagery ‐

SA6 Southern Laos


yes yes yes


Rainfall‐runoff + hydraulic model/satellite imagery Hydraulic model/satellite imagery As above

SA7 Se San/Sre Pok/Se Kong


yes no yes

From rainfall or discharge statistics or regional approach ‐ HYMOS/other databases

Rainfall‐runoff + hydraulic model/satellite imagery ‐ Hydraulic model/satellite imagery

SA8 Kratie Tributary Mainstream Combined

yes yes yes

From rainfall statistics Annual Flood Report No data

Rainfall‐runoff + hydraulic model/satellite imagery Satellite imagery u/s Kratie + model d/s Kratie ‐

SA9 Tonle Sap


yes yes yes

MRC database HYMOS/other databases MRC database

Rainfall‐runoff + hydraulic model/satellite imagery Hydraulic delta model/satellite imagery As above

SA10 Mekong Delta

Tributary Mekong Delta

yes yes

HYMOS/other databases HYMOS/other databases

rainfall‐runoff+ hydraulic model/satellite imagery Hydraulic delta model/satellite imagery



2.3 Tributary flood hazard assessment

2.3.1 General

Flood hazard assessment procedures rely on hydrological monitoring infrastructure in order to provide measurement data, yet such monitoring capacity on the tributaries of the Mekong is currently limited. Only a limited number of basins are equipped with rainfall and water level recorders. In most locations only daily rainfall data are available. The water level gauging network and the monitoring intervals are usually inadequate to properly describe flash floods. Often, only daily water levels are measured and this is unsuitable for flash flood analysis. Flow discharge measurements usually cover only the lower stages and occasionally include flood data. Dependent on data availability one of the following three procedures is feasible for estimating flood hazards:

1. Flood hazard derived from rainfall extremes.

This method is recommended for small basins when little or no historic data is available on flows. It starts with the specification of the meteorological hazard is terms of design rainstorms of desired return periods. The design rainstorms data are adjusted to net rainstorm data and plotted on design discharge hydrographs for the required location using appropriate unit hydrograph and hydrologic routing techniques to arrive at the hydrological hazard measure. Finally, the flood hazard is determined by transforming the peak level data of the discharge hydrographs into water levels by means of flow simulation techniques using a hydraulic model.

2. Flood hazard derived from observed flows. When a homogeneous water discharge data series (of >15 years) is available for a specific river reach, the hydrological hazard can be derived from a statistical analysis of annual flood peaks and flood volumes. A hydraulic model of river and floodplain can then be used to convert the hydrological hazard into a flood hazard. Data from representative flood hydrographs, covering the full spectrum of flood peak and flood volumes, can be converted into water level data using this hydraulic model. Water level and flood volume data can be used in a Monte Carlo procedure to derive the frequency of different water levels as a function of time and space. A Digital Elevation Model (DEM) can then be used to generate flood hazard maps.

3. Flood hazard determined from regional flood statistics. The distribution of annual flood peak can also be obtained from regional flood statistics, when available, if the average annual flood peak is known. For streams without water level recorders and discharge measurement equipment, this value may be estimated from basin characteristics.

2.3.2 Flood hazard derived from rainfall extremes

The method of using rainfall extremes for the assessment of the flood hazard includes the following steps:

1. Collection of relevant information; 2. Development of a database and conduct of field visits; 3. Determination of the design rainstorms for different return periods; 4. Plotting of design rainstorm information on design hydrographs; and 5. Use of design hydrographs to determine design levels. These steps are elaborated below.



Step 1 Collection of relevant information

Data required for flood hazard assessment includes:

1. Digital Elevation Model (DEM) of the area and topographical maps; 2. Soil maps, geological maps and land use maps (past and present); 3. Data on historical floods; 4. Layout of the hydro‐meteorological network in the river basin, its operation and

maintenance; 5. Time series data on rainfall, climate, and, when available, water levels, stage‐discharge

measurements, discharges and sediment of the river basin; 6. Hydraulic infrastructure specifications, dimensions and operation in times of flooding; 7. Survey data of the river (in cross‐sections) and floodplain (DEM) at the reach and upstream

and downstream of the reach under investigation and of relevant hydraulic parameters including river bed composition, vegetation, embankment elevation, etc.;

8. Data on the morphological development of the river and its environs; and 9. Planned land use and infrastructural construction.

Step 2 Database development and field visits

The collected information is next organized in a database consisting of GIS‐related and time series data. In the database it is important to make a clear distinction between original/raw data, data under validation, and fully validated data. There are a number of procedures for preparing data for use in models. For example, rainfall data in weekends is often suspect since there may be no monitoring staff available. Thorough validation, correction and completion procedures should be applied to improve the reliability of the data. Since the flood hazard assessment relies heavily on quality rainfall data, available records have to be thoroughly reviewed to assure their quality. Graphical and tabular comparisons are often useful in preparing the data. In some cases, data from the nearest neighbour (i.e., closest monitoring stations) can be used for cross checking purposes and to fill in missing data. A double mass analysis can also be used. Standard textbooks offer detailed explanations of these procedures. In this data collection and validation process one or more field visits is essential to validate collected information and to assess its quality. Experts should visit key rainfall monitoring stations to review their monitoring practices and. Interviewing residents and officials on the extent of historic floods, land use patterns and construction. The field visits can also be used for collection of additional data or establishment of a (temporary) monitoring program to check on the validity of the data.

Step 3 Determination of the design rainstorms for different return periods

The determination of design storms involves the following steps:

1. Assess the time of concentration (Tc) of the basin up to the location for flood hazard assessment. Time concentration is defined as the travel time required for a drop of water to travel from the uppermost part of the catchment to the outlet. Tc is a measure that expresses the time delay between rainfall and flood runoff.

2. Determine the basin rainfall from point rainfall data for the smallest available interval if applicable.

3. Develop intensity (or depth)‐duration‐frequency curves for specific rainfall return periods of 2 to 100 years, for durations of 5 minutes to 24 hours. If no short interval rainfall data is available, short duration rainfall intensities can be derived from daily rainfall data. In such cases, Extreme Value Type 1 (EV1) or Generalized Extreme Value (GEV)‐distributions can be



fitted to the observed distribution of annual maximum daily rainfall, and the rainfall for selected return periods can be estimated. These values should be adjusted to 24 hr values by applying a correction factor of 1.12. Then regionally acceptable values for the coefficients b and n in the following intensity duration relation are to be obtained:

( )

n

aI

b D (2.1) where: I = rainfall intensity (mm/hour) D = rainfall duration (hours) a,b,n = constants

Values for coefficients for Phnom Penh e.g. are b = 0.23 and n = 0.86 (Watkins et al., 1984).

4. Apply the Rainfall Ratio Method to derive the rainfall for durations shorter than 24 hrs. The Rainfall Ratio method assumes that the constants in equation (2.1) apply for all durations less than 24 hours and all return periods. Thus, the rainfall of duration D for any return period is derived from the 24 hour rainfall extreme via equation (2.1) as follows:

24

24

n

D

D bRR

b D

(2.2) where: RRD = rainfall ratio for conversion of 24 hr rainfall into D hr rainfall

D = rainfall duration (hours) 5. In case of single point rainfall, the point rainfall data needs to be adjusted for the variability

of rainfall over the area (spatial distribution). This can be done through the application of an Areal Reduction Factor (ARF). The ARF is a function of the basin area (A) and rainfall duration (D) and can be derived through statistical processing of single point rainfall data. For example in the case of convective rainfall in Cambodia for durations less than or more than 8 hours and where A is measured in km2,, the ARF is given as follows:

0.33 0.50

0.50

8 : 1 0.04

8 : 1 0.02

for D hrs ARF D A

for D hrs ARF A

(2.3)

6. The design rainstorms for selected return periods can be obtained by the incremental

intensity of the rainfall totals of the calculated durations up to the time of concentration:

1, ,

11

:j jA D A D

j j jj j

R R mmI for D D

D D hr

(2.4) where: Ij = jth incremental intensity (mm/hr)

RA,D = areal rainfall of duration D 7. The incremental intensities can then be arranged around the highest intensity (peak value).

The graphical location of the peak is determined by the storm advancement coefficient r which is the ratio of the time to peak relative to the total duration of the design storm Tc. The advantage of this procedure is that up to Tc all design intensities are included in the design storm. Another option to determine the design rainstorms is by using triangular hyetographs, where the occurrence of the peak value is determined by the storm advancement coefficient. When possible the arrangement is according to observed hyetographs in the region. The availability of short duration rainfall data is then a necessity.

8. Repeating the above step for required return periods yields a set of hyetographs.

MRC Flood

Guidelines

Step 4

The conv1. Dete

ordin2. Deve

haza3. Conv

deve Rainfall ldeterminNumber requires extensive To develmethod. hydrograwith a tithrough When thconvolutperiod eabstractihydrograconditiondifferentthe locat The abovsome mocalculatioallow spcalculatiorainfall‐rsegmentrouting m

Step 5

By meanof requibathymeroughnesinclude pcause fut 2.3.3

To demoRiver neaonly dailfollowing

d Management a

Flood Risk Asse

Using the de

version of desermination ofnates of the helopment of ard assessmevoluting the eloped under

osses are conne rainfall lomethod or Sinformation ely described

lop a unit hyClark's meth

aph and is bame base equa linear reser

he unit hydted to a desequal to theions that caphs under dns assumed t return periotions and to p

ve procedureodification, tons should upatial variations, the basunoff transfot is then addemethods or w

Using design

ns of hydrauliired return etric survey dss. The calcuplanned protture water le

Example of

onstrate howar Pleiku, Viey rainfall serg parameters

and Mitigation P

essment

sign storm to

sign rainstormf the net rainhyetograph; a unit hydrnt is undertanet hyetog

r the previous

nsidered not osses is the SSCS‐CN), appon soil type

d in hydrologi

ydrograph ofhod requires eased on the cual to the corvoir.

rograph has ign hydrograe return periause the ddrier initial cin the compod peak wateplot a frequen

e is applicablethis proceduse a segmenton in the ein up to theormation caned to the arewith the hydra

n hydrograph

c model simuperiods. Th

data on the rulations shoutection worksvels to rise in

flood hazard

w the above et Nam can bries are avails have also be

Programme Com

o determine t

m into a designfall by estim

ograph of thken, and raph to the s activity.

contributingSoil Conservalied to wet as, land use acal textbooks

f the basin itestimation ofconcentrationoncentration

been estabaph, which ciod of the ddifference beconditions, isputation. By er levels, it isncy distributi

e to relativelyre can also bted approachexcess rainfae location of n be applied ea of interesaulic model e

s to determin

ulation, the phe hydraulic iver and floould take into s for investign areas furthe

d derived from

procedure wbe used as aable. For theeen assumed

mponent 2: Struc

‐ 18 ‐

the design hy

gn hydrograpmating rainfal

he basin up

design hyd

g to the floodation Serviceantecedent soand conditions (e.g., Viessm

t is recommef three basin n time Tc, ontime Tc, and

blished, the comprises thdesign rainstetween extrs valid in vierepeating ths possible toon.

y small catchmbe applied th. This will trall and runof interest canto each seg

st. This routinextended to t

ne design lev

peak dischargmodel is

odplain, and taccount exis

gation of alteer upstream.

m rainfall ext

works, a hypon example. Fe developme:

ctural Measures

ydrograph

ph involves thl losses, to b

to the locat

rograph usin

d hazard. Onee Curve Numoil moisture n of the landman, 1989).

ended that uparameters n the routingd on a simula

net design e hydrologictorm. This mremities of ew of the whis process fobtain data

ments, rougho larger areaansform exceoff charactern be dividedgment. The hng can be dothe outlet of t

vels

ges are transfschematised the estimatesting flood pernatives, sin

remes

othetical basFor the calcunt of a desig

and Flood Proof

he following se distributed

ion for whic

ng the unit

e of the methmber methodconditions. Td. The SCS‐CN

use is made ofor the derivag of a time‐aation of chan

rainstorm dal hazard wimethod of restorms and

wet antecedefor design rafor discharg

hly up to 1,00as. For largeess rainfall inristics. In mainto segme

hydraulic outone by simplethe uppermo

formed into waccording

d or calibrateprotection mence these me

in on the Upulations, we agn storm calc

fing

May 2010

steps: d among the

ch the flood

hydrograph

hods used to d (SCS‐Curve This method N method is

of the Clark ation of unit rea diagram nnel storage

data can be ith a return educing the d discharge nt moisture ainstorms of ge values for

00 km2. With r areas, the nto runoff to aking these nts and the tput of each e hydrologic ost segment.

water levels to detailed ed hydraulic easures and easures may

pper Se San assume that culation, the



Basin area of 50 km2;

Rainfall intensity parameters for disaggregation of daily rainfall (equation 2.1) b = 0.23 and n = 0.86 apply;

Storm advancement coefficient r = 0.3;

ARF for convective storms in tropical countries apply; and

Rainfall conditions are similar to those at Pleiku with an orographic adjustment factor of 1.25.

Daily rainfall series data for Pleiku (monitoring station 140703) are available for the years 1927‐2006, though with some gaps. After validation and homogeneity tests, EV1 and GEV distributions have been fitted to the annual maximum rainfall of the monitoring station. The results are shown in Table 2‐2 and in Figure 2‐3.

Table 2‐2 Parameters of EV1 and GEV distributions fitted to annual maximum daily rainfall at Pleiku and

rainfall values (mm) for selected return periods.

Parameters/

Return period (years) Pleiku

EV1 GEV

k

α

u

‐

30.12

98.08

0.0048

30.26

98.11

2

5

10

25

50

100

500

1000

109

143

166

194

216

237

285

306

109

143

166

194

215

236

283

304

Figure 2‐3 EV1 and GEV‐fit to annual maximum daily rainfall at Pleiku for the years 1927‐2006.

Annual maximum daily rainfall at Pleiku, 1927-2006

0

50

100

150

200

250

300

350

-2.00 -1.00 0.00 1.00 2.00 3.00 4.00 5.00 6.00

Reduced EV1 variate

An

nu

al m

ax

imu

m r

ain

fall

(m

m)

Observations

EV1 distribution

GEV distribution



The depth‐duration‐frequency curves have been derived for intervals of 5 min to 24 hrs starting from the daily rainfall extremes for return periods (T) of T = 2 to 100 years, using equation (2.2). The results are presented in Table 2‐3 and shown in Figure 2‐4. The calculation for the T = 50 year design storm is shown in Table 2‐4. The result is presented in Figure 2‐5. It is derived from a re‐ordering of the derived incremental intensities. Here, a basic interval of 15 minutes has been assumed. The sensitivity of the design hydrograph for the choice of this interval should be tested, once the basin Unit Hydrograph is derived. This procedure is to be repeated for other selected return periods that will be used in the flood hazard assessment.

Table 2‐3 Short duration rainfall (mm) at Pleiku derived from daily rainfall extremes.

Duration Rainfall ratio Return Period in years

hrs RR 2 5 10 25 50 100

5 min 0.083 0.15 18 23 27 32 35 39

10 min 0.167 0.24 29 38 44 52 58 63

15 min 0.250 0.30 37 49 56 66 73 81

30 min 0.500 0.42 52 68 79 92 102 112

1 hr 1 0.54 66 87 101 117 131 144

2 hr 2 0.65 79 104 121 141 157 172

4 hr 4 0.75 91 120 139 162 181 198

8 hr 8 0.84 103 135 157 183 204 224

12 hr 12 0.90 110 144 167 196 218 239

18 hr 18 0.96 117 153 178 208 232 254

24 hr 24 1.00 122 160 186 217 242 265

Day Day 109 143 166 194 216 237

Figure 2‐4 Depth‐duration‐frequency curves for Pleiku.

Depth-duration-frequency curves for Pleiku

0

50

100

150

200

250

300

0.01 0.10 1.00 10.00 100.00

Duration in hrs

Ra

infa

ll (

mm

)

T = 2 year

T = 5 year

T = 10 year

T = 25 year

T = 50 year

T = 100 year



Table 2‐4 Computation of design storm ordinates for 15 min intervals.

Time (hrs) 0.25 0.50 1.00 2.00

Point Rainfall (mm) 73 102 131 157

Correction for orography 92 128 164 196

ARF 0.55 0.64 0.72 0.78

Areal rainfall (mm) 51 82 117 152

Incremental intensity (mm/hr) 202 127 70 35

No of 15 min increments 1 1 2 4

Figure 2‐5 A 2‐hour‐design rainstorm diagram for the Upper Se San River derived from measures used for

Pleiku.

In the development of the flood hazard, the next step is the transformation of the design storm into the design hydrograph. This involves, first, the assessment of the excess rainfall. For this, the Curve Number Method is useful. Next, the Unit Hydrograph of the basin is determined for which the Clark Method can be used. This method is based on the routing of a time‐area diagram (with a time base equal to the time of concentration Tc) through a linear reservoir, simulating channel storage. In the diagram of the hypothetical basin, isochrones representing points of equal travel time to the segment outlet (downstream basin limit) can also be constructed. (See Figure 2‐6.) The time area diagram fully represents the physical runoff characteristics from (mountainous) catchments. This diagram can be thought of as the outflow from the catchment if only translation and no deformation takes place of an instantaneous unit supply of rain. Flow paths, slopes and hydraulic roughness determine the speed of water outflow and these characteristics are easy to obtain from a Digital Elevation Model (DEM) or a topographic map. In addition to the concentration time, plotting of the time area diagram requires estimating a water recession coefficient, k, for routing. Note that the output from the reservoir is representative for the instantaneous unit hydrograph (IUH). In order to develop the Unit Hydrograph, the IUH data must be transformed into a Unit Hydrograph (UH) of the required interval by multiplying with

2 hr-design rainstorm for Pleiku, T = 50 years

0

50

100

150

200

250

0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00

Duration (hrs)

Ra

infa

ll (m

m/h

r)



observed rainfall depths. The two parameters, Tc and k, can also be obtained from observed rainfall and discharge hydrographs.

Figure 2‐6 Principles of the Clark Unit Hydrograph method, shown in various diagrams that are related to

each other.

The time of concentration is equal to the time interval between cessation of rainfall and the time the hydrograph has receded to its inflection point. Note that this method is only applicable when drainage conditions and infrastructure are constant and are not expected to change. If conditions are not constant, the time of concentration can be determined from physical features of the catchment segment such as length and slope. The recession coefficient, k, can be calculated from the surface runoff hydrograph by measuring the slope of the curve in the area where it descends (the area of flood water recession). The Unit Hydrograph can be used to generate the excess hyetograph and the data can then be used to calculate the design flood for the selected return periods. Using these results, it is possible to generate a frequency distribution of the floods resulting from the selected rainfall duration (equal to the time of concentration). It is to be verified whether storms of shorter or longer duration lead to higher peak discharges. The height of the peak discharge relative to storm duration is dependent on the intensity distribution in the rainstorm and the way the losses are distributed over the rainfall ordinates. The data for frequency distribution of flood peak can be transformed to design flood levels through calculations using measurements of backwater. If upstream water storages are significant, then the full flood hydrograph should take the entire hydraulic network into account, using realistic initial values for the storage capacities.

area

k

Tc

UH

(t)

IUH UH

time

Time-area diagram

Linearreservoir

1-hourUnit

Hydrograph

1

23

45

67

isochrones

segment

7 6 5 4 3 2 1

time



2.3.4 Flood hazard determined from observed flows

When a water discharge data series of sufficient length (≥ 15 years) is available, the hydrological hazard can be determined by statistical analysis of these data, provided that the time series data are reliable and homogeneous. This procedure involves the following steps:

1. Collection of relevant information; 2. Database development and field visits; 3. Assessment of the hydrological hazard in terms of flood peak water levels and flood

volumes; 4. Conversion of selected flood hydrograph data to flood level data using a hydraulic

model; 5. Flood hazard assessment by means of a Monte Carlo sampling procedure, interpolating

between the simulated water levels based on samples from the flood peak water levels and volume distributions; and

6. Preparation of flood maps by selection of flood levels, durations etc. of the same return period.

Step 1 Collection of relevant information

The required data for this type of flood hazard assessment does not differ much from the one above. (See Step 1 in the procedure described in Section 2.3.2.) Here, emphasis is on several factors in the hydrometric (water measurement) network layout and operations. These include sampling and measurement procedures, monitoring station and equipment maintenance, gauge ranges and re‐settings, available water level data series, stage‐discharge data and assumed discharge ratings with extrapolation procedures and discharge series data.

Step 2 Database development and field visits

(See also Step 2 in the procedure described in Section 2.3.2.) Thorough validation, correction and completion procedures (filling in missing data) should be applied to improve the reliability of the water discharge data. This requires inspection of all hydrometric monitoring stations relevant to assessment of the hydrological hazard and validation of the data from key stations. Flood marks should be levelled and discussed with local residents. Sometimes there is lateral water inflow. In such cases, when rainfall data are used to determine the lateral inflow, the rainfall network should be inspected as part of the validation procedure. The hydraulic infrastructure and flood protection works should also be inspected on field visits, to ascertain their capacity and operations in times of floods and to determine their impact on the shape of the hydrograph curves as a result of water storage and/or backwater. Construction in the basin, which can affect the hydrograph (specifically the peak value), should be quantified. Screening of historical water levels involves inspection of the historical patterns of hydrographs, comparison with hydrographs at nearby monitoring stations, stage relation curve analysis to identify outliers and shifts in gauge locations, and comparison of hydrographs with hyetographs to detect anomalies in the rainfall – runoff process. Special attention should be paid to differences between instantaneous peak water levels and maximum daily average water levels. The latter information is generally only available in the databases. The peak values of the hydrographs should later be adjusted in case of differences between instantaneous peak water levels and maximum daily average water levels. Stage relation curves describing the relation between water levels and discharges at gauging stations may be applied to fill in gaps in the data of a station (with appropriate time shifts to account for flood wave celerity). With respect to the stage‐discharge data the stage data is first



compared with the stage‐hydrograph. The computation of the discharge from field data is checked and its accuracy is assessed. Possible backwater and unsteady flow effects are to be identified by means of a first order backwater calculation and assessment of the Jones correction, respectively. The establishment of the discharge ratings is to be inspected, with special emphasis on the extrapolation. The latter should be physically based taking into account flood protection works in and downstream of the river section. The next step is to apply the discharge rating to the water level data series in order to compute the discharge data series. A comparison is made with the series in the database. Discharge hydrographs at neighbouring monitoring stations can be compared graphically and the consistency of the data can be verified by means of water balances and double mass analysis. Similar comparisons with data at neighbouring monitoring stations can be used to validate peak values. Finally, peak values should be examined, graphically, statistically and using documents on drainage conditions to spot possible trends indicating problems in the data and other anomalies. Data should be corrected to eliminate these trends and anomalies. Special attention should be paid to the variability of the discharge ratings at locations where there have been morphological changes in the river bed. The variation in water levels for particularly high discharges over the years should be calculated and then applied to assessing the additional uncertainty in water levels for the specific discharge rating used in the model.

Step 3 Assessment of the hydrological hazard

The features of the hydrograph which are of importance for flood hazard assessment include the peak value considering maximum water levels and the flood volume. The flood volume determines the duration of flooding. The following tasks are required in the assessment process: 1. For flood peak water levels:

Creation of a homogeneous series of annual maximum flows of sufficient length;

Application of extreme value analysis to the annual maximum flow series to derive frequency distributions for the flow extremes;

For locations upstream of Vientiane it is advised using comparisons of the obtained distribution with the regional flood frequency curve described in Section 2.3.5.

2. For flood volumes:

Selection of data from 20 to 25 of the largest recorded floods;

Derivation of a regression line that shows the relationship between flood peak discharge (Q) and flood volumes (V) as V = f(Q) and its scatter (the deviation from the fitted line), or “standard error”, (Se). The regression line gives the “average” or “expected” value of the flood volume, for given values of the flood peak. The expected flood volume increases in relation to the increasing flood peak. Note that the relationship is not perfectly described by the fitted line. There is some scatter (volumes above and volumes below the expected values of the fitted line). Figure 2‐7 offers an example of such a regression line and the deviation from the line for a single point. The value, Se, is the standard deviation of all points with respect to the regression line. It can be used (in step 3) to take the effect of the scatter into account;

Derivation of representative dimensionless flood hydrographs from the selected hydrographs. The base of the hydrograph should be long enough to represent the full duration of characteristic floods (from M days before until N days after the peak water level). For each flood assign the time of Qmax (the peak discharge of the flood) as the initial time (t) such that t=0. Then, determine for the range of times (t = ‐M, …, t=0, ..., t=N) the water discharge for each time relative to Qmax (=Q/Qmax). Repeat the procedure for all other selected peak discharges (for the different return period floods).



Then, apply a frequency analysis on Q/Qmax for each time measurement (t = ‐M, …, t=0, ..., t=N). The result will be a distribution function of Q/Qmax for each time at which measurements are taken. Thus, for any give probability, p, of (non‐)exceedance the value of Q/Qmax will be known for all time at which measurements are taken. Connection of these values yields the (dimensionless) hydrograph for p. In this manner, for a variety of p‐values, hydrographs can be generated for use in the next step.

3. For creation of a flood peak water level‐flood volume matrix for use in hydraulic simulations:

Selection of flood peaks of return period 2, 5, 10, 25, 50 and 100 years;

Determination of a range of flood volumes for the selected flood peak values, based on the scatter plot data around the flood volume (V) ‐ flood peak discharge (Q) regression line relationship, V = f(Q). For each selected peak value (the different return period floods), we can generate a very low (R‐1.96 Se), a low (R‐Se), medium (R), high (R+Se) and very high flood volumes (R+1.96 Se) – R being the mean of the flood volumes ‐ by scaling of the representative flood hydrographs to the flood peak value of selected return periods. This results in 30 (6 return periods x 5 flood volumes = 30) hydrographs;

Adjustment of flood hydrographs for differences between daily average flow maxima and instantaneous peak flows;

Lateral water inflows, assumed to be relatively small, can be defined as a percentage of the selected flood hydrograph measurement.

Figure 2‐7 Hypothetical scatter plot of data points and fitted regression line showing the relationship

between flood volumes and flood peak water levels. The red arrow in the Figure shows the

deviation from the regression line for a single observation.

Step 4 Conversion of flood hydrographs to flood levels using a hydraulic model

There are a number of options of hydraulic models for use in converting flood water data into flood levels. An accurate hydraulic model is needed for the river and floodplain being analysed. The model’s boundaries should be chosen in accordance with the hydrometric network equipment and the boundaries of the area of analysis. It is suggested using a one‐dimensional (1D) – two‐dimensional (2D) hydraulic model, where rivers are represented as 1D and the floodplain as 2D, both derived from a DEM. This approach has clear advantages over a quasi‐2D approach, which is a 1D model with storage cells for the floodplain, where the need to make

regression flood peak/flood volume

0

500

1000

1500

2000

2500

0 20 40 60 80 100 120 140 160 180

flood peak

floo

d vo

lum

e

deviation from regression line

regression line



qualitative assumptions or estimates on the connection between the river and floodplain introduces subjectivity. The 1D‐2D model has also advantages over a full 2D model, since narrow rivers cannot be properly represented in the 2D model. Any selected model should be calibrated for river and floodplain conditions representing the base case. Special attention should also be given to the position of the river bed in comparison with possible river bed shifts due to natural or human induced morphological developments. The 30 selected hydrographs are subsequently routed in the model through the river and floodplain system, and the water levels are computed as functions of time and space which are placed in a database for use in a Monte Carlo simulation procedure for flood hazard assessment. These simulations can be repeated to test the results of construction plans that are under consideration, after adjustment of the hydraulic model for the different scenarios.

Step 5 Flood hazard assessment

The probability of exceedance of a water level at any location can be determined using Monte Carlo simulations. Monte Carlo simulations of water level exceedance use a large number (N) of paired samples of the peak discharge and flow volume. In this procedure, the derived frequency distribution of peak water level discharges is used to sample the peak discharges. Subsequently the flood water volume can be derived from the peak discharge, taking the sum of: 1. the regression line to derive the “expected” volume, given the value of the peak water level

discharge; and 2. a sample from the normal distribution with mean 0 and standard deviation Se (see step 2)

to account for deviation from the expected volume. Application of this procedure ensures that sampled volumes are correlated with the sampled peak discharges to reflect the relationship in the raw data. For each sampled pair of water volumes and peak discharges, the water levels at the areas under study are derived from 2‐dimensional interpolations of the grids of the 6 x 5 simulated hydrographs described above. The result is an N‐year synthetic series of water levels (h1, …, hN,) at each location. From this series the T‐year return period flood water levels can be easily derived with an equation such as Gringorten’s formula:

0.44

0.12i

i

rp

N

(2.5) where: pi = probability of exceedance of synthetic water levels of height hi for I,

the particular area of interest under study; N = total number of Monte Carlo samples;

(1 = highest, N = lowest), and ri = rank number of the maximum synthetic water levels hi

Step 6 Preparation of flood maps

The mapping of the hydraulic modelling results is described for mainstream flows in Section 2.7. For tributary floods the mapping principles are the same as those for mainstream flows, but the data interpolation is more critical because of the hilly terrain. 2.3.5 Flood hazard derived from regional flood statistics

Adamson (2007) proposed the use of a regional approach to flood hazard assessment. The method has been adapted for use in the upper part of the LMB for 16 drainage basins with



areas ranging from 200 to 6,000 km2 and for 1 basin of 19,700 km2. The technique is applicable provided the regional curve for the upper part of the LMB is valid for the particular basin. Its use involves the following steps:

1. Creation of a regional sample of annual maximum flood peak water levels by pooling together the individual annual maximum values, scaled to their individual mean annual flood value. Discordance and homogeneity tests should be applied to assure regional homogeneity of data from the selected monitoring stations (see Hosking and Wallis, 1997).

2. Fitting of an extreme value distribution to the scaled ranked annual extremes. For the LMB basins considered by Adamson (2007), the Two Component Extreme Value (TCEV) distribution fits the observed frequency distribution of these pooled values. The regional curve for the upper part of the LMB is shown in Figure 2‐8. The TCEV‐distribution was chosen to account for different phenomena creating the water discharge extremes (i.e. monsoon and typhoons). Typhoons create extremes far beyond the monsoon levels.

3. Applying the TCEV‐values to un‐gauged sites. A method is applied using regional data for the mean annual maximum floodwater discharge that describes the relationships between floodwater discharge and one or more climatic and/or basin characteristic(s). This relationship allows for floodwater discharge rates to be determined anywhere in the region if those other characteristics are measured. For the upper part of the LMB, the following relationship between floodwater levels and drainage area was identified:

0.75 .aveQ A err (2.6)

where: Qave = mean annual maximum floodwater discharge rate (m3/s) A = drainage surface area (km2)

The regression error (dispersion of the data around the fitted line) is considerable (Se is about 50% of Qave), suggesting that additional descriptive variables characterising the area are needed to fully describe the floodwater discharge rate.

Figure 2‐8 TCEV‐distribution of the ratio of the T‐year return period event to the mean annual flood for the

pooled regional sample (Upper part of LMB, from Adamson, 2007).

Estimated TCEV distribution of the ratio (%) of the T-year event to the mean annual flood for pooled regional sample (n=530) (Source: Adamson (2007))

0

100

200

300

400

500

600

700

1 10 100 1000

Return Period (years)

Rati

o (

%)

of

T-y

ear

flo

od

to

mean

an

nu

al fl

oo

d Lower 95% CL

TCEV-values

Upper 95% CL



In using the regional approach, the following remarks are made:

In principle the method is very attractive and easy to use, provided that the base data can be properly validated and a clear relationship can be established between mean annual floodwater discharges and meteorological and river basin characteristics.

The data set used in Adamson (2007) has been taken from the MRC database. Inspection of the data reveals inconsistencies in the data for flow extremes across different monitoring stations.

Using Adamson’s data set, Qave appears to correlate well with the mean annual floodwater flow. This, in turn should be directly related to annual rainfall which are generally available for un‐gauged streams in view of the better distribution of rain gauges. Furthermore, other data can be used in other equations that predict maximum floodwater discharge rate or rainfall (such as stream length, basin slopes and land use data from available DEM), to improve the estimate of Qave values.

An adjustment is required for instantaneous peak values relative to daily average maximum values in the above approach. In the case of flash floods a considerable adjustment is expected as the flood may last only for a number of hours.

2.4 Mainstream flood hazard assessment

2.4.1 General

The hydrological hazard along the Mekong River is determined by the discharge peaks on the river that generate the maximum water levels and by the flood volumes determining the duration of flooding. Conversion of hydrological hazard data into flood hazard data requires using a hydraulic model of the river and floodplain to convert water flow data into water level data. Data from selected hydrographs covering the full spectrum of possible combinations of flood peaks and of flood volumes can be used in the model. Special attention should be paid to variations of the river bed due to natural or human induced morphological changes, which affect the water level discharge ratings along the river. The Monte Carlo simulation sampling technique can be used to derive exceedance probabilities of water levels and floodwater damage. The simulation generates results for a large number of combinations of peak discharges, flood volumes and river bed conditions and this data can be used to estimate floodwater damage.

The flood hazard assessment for mainstream floods includes the following steps:

1. Data collection; 2. Database development and conduct of field visits; 3. Data validation and processing; 4. Hydrological hazard assessment in terms of flood peaks and flood volumes; 5. Flood hazard assessment using a hydraulic model that transforms selected flood

hydrograph data to flood level data and applies Monte Carlo sampling to determine flood levels and durations for distinct return periods;

6. Preparation of flood maps through selection of flood levels, durations, and other features of the same return periods. (See Section 2.7).

The steps are elaborated in the following sub‐sections. 2.4.2 Data collection

For assessment of the hazard of mainstream floods, the following data and information are required:



1. Identification of key hydrological monitoring stations (Qh‐stations) along the Mekong River in the area of study;

2. Survey data of the Mekong River (full river bathymetry) and of the floodplain in the study area and beyond and relevant hydraulic parameters including river bed material, vegetation, embankment elevation, etc.;

3. Updated Digital Elevation Model (DEM) of the floodplain in the area under study; 4. Soil maps of the floodplain; 5. Land use maps of the floodplain, including past, present and expected future conditions; 6. Data on hydraulic infrastructure in and downstream of the study area, including its

dimensions and operation in times of flood and planned future developments; 7. Historical flood maps (including flood levels, extent, depth and duration); 8. Layout, operation and maintenance information for the hydro‐meteorological network

around the study area including information on the nearest upstream and downstream monitoring stations along the Mekong River and monitoring stations on the tributaries where relevant;

9. Time series data of water levels, stage‐discharge measurements, discharge ratings and discharges series of the relevant monitoring stations as well as series data of rainfall and climatic variables in the region;

10. Changes that have occurred in runoff characteristics of basins upstream of the study area that affect the homogeneity of the historical water discharge data series (including construction of reservoirs and land use changes).

Below, is a list of the agencies in the MRC Member Countries that are responsible for collecting and providing this data. Cambodia Ministry of Water Resources and Meteorology (MOWRAM)

Department of Hydrology and River Works (DHRW), for water level, discharge and rainfall time series;

Department of Meteorology (DOM), for rainfall, climatic data (air temperature, relative humidity, wind direction, wind velocity, duration of sunshine, radiation)

Ministry of Public Works and Transport (MPWT)

Department of Waterways River and Lakes for bathymetric data, hydrographic atlases;

Department of Road and Transport Ministry of Agriculture, Forestry and Fishery (MAFF)

Department of Agronomy and Agricultural Land Improvement for soil maps and land Use maps)

Phnom Penh Municipality:

Department of Public Works and Transport of Phnom Penh Municipality for urban flood protection, dykes/drainage in and around Phnom Penh

Ministry of Land Urban Planning Land Use and Construction (MUPLC)

Department of Land Use Planning, Commune Land Use Planning (CLUP) Lao PDR Water Resources and Environment Administration (WREA)

The WREA is a new government agency that combines what were previously the Department of Meteorology and Hydrology (DMH) of the Ministry of Agriculture and the Waterways Administration Division (WAD) of the Ministry of Communication, Transport,



Post and Communication. WREA is responsible for collecting and providing all related water data in Lao PDR as well as for the establishment and management of the country’s hydrological and meteorological network. The Hydrographic Atlas data of the Mekong River mainstream has also been transferred to the WREA.

Thailand Ministry of Agriculture and Cooperatives (MOAC)

Royal Department of Irrigation (RID), for hydro‐meteorological data in most of the Mekong River tributaries, especially the Nam Mae Kok, Nam Mae Ing, Nam Chi and Nam Mum Rivers.

Ministry of Natural Resources and Environment (MNRE)

Department of Water Resources for management of hydro‐meteorological networks (previously under the Department of Energy Development and Promotion), including all the Thai Mekong River mainstream monitoring stations. Climatic and water quality data collection previously supported by the Mekong Committee were also collected and provided by this department including bathymetric survey data and hydrographic atlases.

Ministry of Information and Communication Technology (MICT)

Department of Meteorology (TMD), for meteorological data, climatic data, and weather forecasting.

Ministry of Interior (MOI)

Department of Disaster Prevention and Mitigation, Department of, Local Administration, Department of Public Works, for urban and rural planning and flood mapping.

Viet Nam Ministry of Natural Resources and Environment (MoNRE)

Department of Water Resources Management

Hydro‐meteorological Services of Viet Nam, Southern Region Hydro‐meteorological Center for hydrological, meteorological, water quality, and salinity data collection network management, data collection and dissemination, and flood, salinity and weather forecasting data.

Sub National Institute for Agricultural Planning and Production (NIAPP) for flood maps in the Mekong Delta and soil suitability maps.

Southern Institute of Water Resources Research (SIWRR), for hydraulic works (sluices, hydropower, reservoirs, sea, dykes, embankments), river training works, coastal zone management (river morphology, strategies and work for river bank protection), irrigation, drainage, soil reclamation, water supply and drain information. The SIWRR also operates and maintains meteorological and hydrological data, and cross‐section data for the Mekong Delta and Dong Nai River basin.

2.4.3 Database development and conduct of field visits

The collected information is to be organised in a database consisting of GIS‐related and time oriented data. In the database a clear distinction is to be made between original/raw data, data that is being validated, and fully validated data. During the field visit the data on flood protection and hydraulic infrastructure in the study area should be checked, including maintenance and operation in times of flood. All key hydrometric monitoring stations should be visited and monitoring and discharge measurement practices



should be reviewed. Gauge stations and gauge histories should be inspected, including gauge reach, staff gauge connections, bench marks, gauge zero controls and shifts, gauge damages and repairs as well as records from comparisons of manual and automatic gauges. Flow measurement equipment calibration and functioning should be tested. Residents and officials should be interviewed on the extent of historic floods, flood protection developments and land use changes. Flood marks should be levelled and discussed with local residents. In connection with the field visit and where required, additional data can be collected or a (temporary) monitoring program can be established and implemented. 2.4.4 Data validation and processing

Thorough validation, correction and completion procedures (filling in missing data) should be applied to improve the reliability of the water discharge data (See also Section 2.3.2). The process should include three components:

Screening of historical water levels involves inspection of the historical patterns of hydrographs, comparison of hydrographs at nearby monitoring stations, stage relation curve analysis to identify outliers and gauge shifts. Stage relation curves (with appropriate time shifts to account for flood wave celerity) may be applied to fill in gaps in the data.

Stages of discharge measurement data should first be compared with the stage hydrograph. The computation of the discharge from field data is checked and its accuracy is assessed. Possible backwater and unsteady flow effects are to be identified by means of a first order backwater calculation and assessment of the Jones correction, respectively. The establishment of the discharge ratings is to be inspected, with special emphasis on the extrapolation. The latter should be physically based, taking into account flood protection works in and downstream of the section. The variability of the discharge ratings at locations as a result of natural or human induced morphological changes in the river bed can be assessed, by determining the variation of the water level for distinct high water discharges through the years. Particularly, this change is assessed for the natural bank full floodwater discharge.

The next step is to apply the discharge rating to the water level data series in order to compute the water discharge data series. A comparison is made with the series in the database. Discharge hydrographs at neighbouring monitoring stations can be compared graphically and the consistency of the data at each hydrograph can be checked by means of water balances and double mass analysis. Comparisons with data at neighbouring monitoring stations can be used to validate peak water level measurements. Finally, peak values should be examined, graphically, statistically and using documents on drainage and storage conditions to spot possible trends indicating problems in the data and other anomalies. Data should be corrected to eliminate these trends and anomalies.

2.4.5 Hydrological hazard assessment in terms of flood peak water levels and flood volumes

Hydrological hazard is determined by peak flow and flood volume and their interaction. A matrix of flood hydrographs can be generated covering the full spectrum of possible hydrographs with peak water level values for return periods of more than or equal to two years (T ≥ 2 years). The development of this matrix involves the following tasks: 1. Creation of homogeneous water discharge series data for the key measurement station at

the upper boundary of the river reach under study; 2. Estimation of the marginal distributions of annual flood peak water levels and flood

volumes (between June 1st and November 30th) at this upper boundary. The observed distributions of flood peak water levels and flood volumes are generally represented effectively by a GEV distribution;



3. Development of a regression equation that describes the relationship between flood volumes (V) and flood peak water levels (Qp) V = f(Qp) and estimate its standard error SeV, in a process similar to step 3 in Section 2.3.4;

4. Specification of the matrix of flood hydrographs for distinct discharge peak water levels, similar to step 3 in Section 2.3.4.

Changes in the stage‐discharge rating due to river bed variations can be taken into account by changing the hydraulic roughness used in the hydraulic model commensurate with the change in the water level. For this purpose, one should start by assessing the annual changes in the water level for natural bank full flow. The effect of roughness can be determined from comparisons of applied discharge ratings or plots of discharge data series versus water level data series, (provided that shifts in gauge zeros have been eliminated). The differences are given relative to the levels from the discharge rating used in the calibration of the hydraulic model (see Section 2.3.6) and are determined for all key monitoring stations in the modelled river reach. According to Manning’s equation for a wide cross‐section, if the water depth at bank full discharge changes from the original depth, h1, to a second depth, h2 ,with h2 = h1 + Δh, there is a simple equation for the hydraulic roughness, n, over the same cross sections. The new hydraulic roughness value is described by the relationship:

5 / 3

22 1

1

hn n

h

(2.7)

Note that the stage‐discharge relation at the downstream boundary has to be adjusted to avoid unwanted backwater effects. 2.4.6 Flood hazard assessment

The hydrological hazard in terms of peak flow statistics can be converted into water level data by means of a hydraulic model of the concerned Mekong River reach. An accurate hydraulic model is needed for the river and floodplain being analysed. The model’s boundaries should be chosen in accordance with the hydrometric network equipment and the limits of the area of analysis. It is strongly advised using a 1D‐2D hydraulic model, where the Mekong River is represented as 1D and the floodplain as 2D, both readily derived from a DEM. This approach has clear advantages over a quasi‐2D approach with storage cells, where qualitative judgments on the connection between river and floodplain introduce subjectivity. The chosen model should be calibrated for river and floodplain conditions representing the base case. Special attention should also be given to the position of the river bed (particularly in case of possible river bed shifts due to natural or human induced morphological changes). Data from the 30 selected hydrographs can then be input using the model for the particular river and floodplain system. The water levels that are computed as functions of time and space can be placed in a database. The procedure can be repeated at least twice with adjusted hydraulic roughness and downstream stage‐floodwater discharge relations to cover the full range of water level changes for bank full floodwater discharge. These results can be added to the database before using a Monte Carlo simulation procedure for flood hazard assessment. These simulations can be repeated to test the results of construction plans that are under consideration, after adjustment of the hydraulic model for the different scenarios. The final step is the execution of a Monte Carlo simulation as described in step 5 of Section 2.3.4.

MRC Flood Ma

Guidelines Flo

2.4.7 E

Preparation

Figure 2‐9 discharges derive the years (Tablevolumes frofitted regreregression l+Se, +1.96 SThen, for evolumes ca

relationshipfitted lines For each ofvarious locaMonte Cardischarge a(i.e. peak dhydrographhydrographselection pthen “scalethe associapeak dischato decreasresembles t

Figure 2‐9

anagement and

ood Risk Assessm

xample: Mek

ns for Monte

shows the Gof the Mekofloodwater de 2‐5). Subseom June 1st uession equatiline is also deSe from the reach selectedan be speci

p. Figure 2‐11from Figure 2

f the 30 grid ations for a rrlo analysis. as input serieischarge andhs, observed hs that most rocedure is dd” in such a wated grid poinarge was slige the flow the observed

Fit of the GE

Phanom.

Mitigation Prog

ment

kong near Na

e Carlo analys

GEV‐function ong River at discharges coequently, the until Novembon shown asepicted in Figegression lind peak dischafied accordi

1 shows the 2‐10. (This re

points a modrange of hydrFor these s

es data. Thesflow volumehydrograph

closely resemdepicted by tway that its pnt. Figure 2‐htly altered, volume. The hydrograph.

V function to a

gramme Compon

khon Phanom

sis

fitted to thNakhon Pha

orrespondingrelation betw

ber 30th) can s the full linegure 2‐10 by te (with Se tharge‐flood voing to the

resulting 30 epeats some o

del simulatioraulic conditisimulations, se (synthetic)e) as the 30 ghs can be semble the featthe green linepeak discharg13 shows anbut the reme resulting .

nnual maximum

nent 2: Structur

‐ 33 ‐

m (Thailand)

he series of oanom, Thaila to return peween peak flbe derived fre in Figure 2‐the 4 dotted e standard dolume in Tabspread in t

(65) grid pof the proces

n with ISIS caions. The simISIS requires) hydrographgrid points. Inelected and tures of the ges in Figure 2ge and flow vn example ofmainder of thehydrograph

m daily discharg

al Measures and

observed annd. This funeriods of 2, 5ows and floorom a regress10). The sprelines at distaneviation fromble 2‐5, the 5he flood vo

oints in comses as shown

an be run to mulations servs 30 hydrogs need to han order to obt(slightly) ad

grid points sh2‐12). The obvolume are exsuch scalinge hydrographis a realistic

ge of the Meko

d Flood Proofing

nual maximunction can be5, 10, 25, 50od volumes (tsion analysis ead around tnces of ‐1.96m the regress5 accompanyolume‐peak

bination withn in Section 2

derive waterve as the basgraphs of daave the sametain realistic dapted. The hould be selebserved hydrxactly equal g. In this exah was slightlyc model tha

ong River at Nak

g

May 2010

um water e used to 0 and 100 total flow (with the this fitted 6 Se, ‐Se, 0, sion line). ying flood discharge

h the five .3.3.)

r levels at sis for the ily water e features synthetic observed cted. This rograph is to that of ample the y lowered at closely

khon

MRC Flood

Guidelines

Table 2‐5

pa

return p

Figure 2‐1

d Management a

Flood Risk Asse

Param

River at N

arameter

k

α

u

period (years)

2

5

10

25

50

100

0 Peak wat

and Mitigation P

essment

meters of the G

Nakhon Phanom

) discha

er level dischar

Programme Com

GEV distribution

m and discharge

value

0.309

4,685

24,475

arge (m3/s)

26,098

30,097

32,070

33,989

35,090

35,970

rge‐Flood volum

mponent 2: Struc

‐ 34 ‐

n fitted to annu

es (m3/s) for se

me relation at N

ctural Measures

ual maximum da

lected return p

Nakhon Phanom

and Flood Proof

aily discharge o

eriods.

m.

fing

May 2010

of the Mekong



Figure 2‐11 Defined 2D grid of discharges (with return periods 2, 5, 10, 25, 50 and 100 years) and volumes

(based on the regression line and lines of deviation from the regression line).

Figure 2‐12 Observed combinations of peak discharge and flow volume (red dots) and selected 2D grid (blue

dots). The green lines show the observed combinations that most closely reflect the grid

combinations.

25 30 35 40 45140

160

180

200

220

240

260

280

300

320

340

V=f1(Q)-1.96Se

V=f1(Q)-1.00Se

V=f1(Q)

V=f1(Q)+1.00Se

V=f1(Q)+1.96Se

peak discharge [1000m3/s]

volu

me

[BC

M]

station: Nhakon Phanom; selected grid

10 15 20 25 30 35 40100

150

200

250

300

350

peak discharge [1000m3/s]

volu

me

[BC

M]

station: Nhakon Phanom; selected grid and observed combinations

observed

grid



Figure 2‐13 Example of scaling of an observed hydrograph to make it suitable for serving as one of the grid

points in Figure 2‐12.

Monte Carlo analysis and results

For this example, a Monte Carlo analysis was executed with N=100,000 samples. For the resulting N water levels the probability of exceedance was estimated through application of the formula (2.5) presented in Section 2.3.4. For a floodplain just downstream of Nakhon Phanom, the resulting frequency curve for the water level is presented in Figure 2‐14. It starts with a basic level (measured in meters over Mean Sea Level (MSL) of just over 130 m+MSL. This is the output of the ISIS‐model if the floodplain location does not flood during the year. The probability that this happens is approximately 0.5 per year. Naturally, for increasing water levels the probabilities of exceedance are lower. The once every 100‐year floodwater level is approximately equal to 134.8 m+MSL (where the graph crosses the line of x = 10‐2 per year). Similarly, the once every 100‐year flood water level can be determined for other locations. Based on the results for all locations a spatial map can be produced using the GIS‐procedures, described in Section 2.7.

14/05 03/06 23/06 13/07 02/08 22/08 11/09 01/10 21/10 10/11 30/110

5

10

15

20

25

30

date

disc

harg

e [1

000m

3 /s]

(scaled) hydrograph at station Nhakon Phanom; year: 1972observed peak: 25500 [m3/s]; scaled peak: 26098 [m3/s]observed volume: 172 [BCM]; scaled volume: 165 [BCM]

observed

scaled

MRC Flood Ma

Guidelines Flo

Figure 2‐14

2.5 C

2.5.1 G

Combined fis determinRiver, backi The hydrolothe Mekongwater can bnear the rivcombinatiomodel shouan area witarea of themonitoring discharge re The simuladerive the usually inclhigh water floodwaters

anagement and

ood Risk Assessm

Derived freq

floodplain ju

Combined flo

General

floods are floned by the coing up into th

ogical hazardg River can bbe convertedver mouth usons of upstreuld encompah no backwate Mekong Rstations. Th

elation down

tion results exceedance udes the folllevels in thes:

Mitigation Prog

ment

uency curve fo

ust downstream

ood hazard a

oods that occuombination ohe tributary a

of tributary be simulated into water leing this modeeam floodwass the full arter flows andiver around he model apnstream.

derived fromprobabilitiesowing three e downstream

gramme Compon

or maximum an

m of Nakhon Ph

assessment

ur in the dowof tributary wand impeding

runoff waterwith a hydraevels in riverel. The modeater inflow area includingd extending tothis conflue

pplies discha

m this models of water leor four random part of the

nent 2: Structur

‐ 37 ‐

nual water leve

anom.

wnstream secwater flow ag drainage.

rs with high waulic model. r and floodplael uses data aand downstr the tributaro the area whence where trge hydrogra

l can be useevels and flooom variablese tributary th

al Measures and

els at a location

tions of tribund the water

water levels aThe levels ofain in the bacacross the fulream water y and its floohere the rivethere shouldaph data up

d in a Montodwater dam, representinhat is backed

d Flood Proofing

n in the Mekong

utaries. The fr level in the

at the confluef river tributackwater affecl spectrum olevels. The odplains starrs meet, as wd be key hydpstream and

te Carlo procmages. The png the main cd up with ma

g

May 2010

g River

lood level e Mekong

ence with ary runoff cted zone f possible hydraulic rting from well as the drometric a stage‐

cedure to procedure causes for ainstream



1. maximum water discharge volumes in the Mekong River at the confluence of the Mekong and a tributary;

2. total volume of the floodwater in the Mekong at the confluence; 3. total volume of the tributary floodwaters; and 4. maximum water discharge volumes in the tributary. Note that it may not always be necessary to explicitly take the tributary peak flow into account, particularly when floodplain inundations and associated damages are considered. In such cases the tributary flood volume, rather than the peak flow plays a more important role. Tributary peak flow is mainly a concern in the design of flood protection works along the tributary, in which case tributary flood volumes are not the relevant concern. In practice, generally only one of the two is considered and rarely both. Special attention should be paid to variations of the river bed as a result of natural or human induced morphological changes. These can affect the conveyance of the rivers and hence the resulting stages. The river conveyance may be an additional variable that is needed in the Monte Carlo simulation procedure. Finally, combining the water levels of equal probability and comparing the results with a DEM, flood maps of levels, depth and duration can be produced. The flood hazard assessment for combined floods includes the following steps:

1. Data collection; 2. Database development and conduct of field visits; 3. Data validation and processing; 4. Hydrological hazard assessment in terms of flood peak water levels and flood volumes on

the Mekong River and the specific tributary. 5. Flood hazard assessment using a hydraulic model that transforms flood hydrograph data to

flood level data and interpolates from simulated water levels by means of a Monte Carlo procedure to arrive at flood levels and durations for distinct return periods.

6. Preparation of flood maps (see Section 2.7).

The steps are elaborated in the following sub‐sections. 2.5.2 Data collection

For assessment of the hazard of combined floods, the following data and information are required:

1. Identification of key hydrological monitoring stations (Qh‐stations) on the Mekong River tributary extending beyond the backwater reach of the mainstream and the along the Mekong River in the area of study;

2. Survey data of the tributary and Mekong River (full river bathymetry) and of floodplains in the study area and beyond, and data of relevant hydraulic parameters, including river bed material, vegetation, embankment elevation, etc.;

3. Updated Digital Elevation Model (DEM) of the floodplain in the area under study; 4. Soil maps of the floodplains; 5. Land use maps of the floodplain, including past, present and expected future land use; 6. Data on hydraulic infrastructure in and downstream of the study area, including its

dimensions and operation in times of flood and planned future developments; 7. Historical flood maps (including flood levels, extent, depth and duration); 8. Layout, operation and maintenance information for the hydro‐meteorological network

around the study area including information on the nearest upstream (and for the Mekong



River, downstream) monitoring stations along the tributaries and Mekong River where relevant;

9. Time series data of water levels, stage‐discharge measurements, discharge ratings and discharges series of the relevant monitoring stations as well as series data of rainfall and climatic variables in the region, particularly in cases when lateral water inflow downstream of the key monitoring station on the tributary is substantial;

10. Changes that have occurred in runoff characteristics of basins upstream of the study area that affect the homogeneity of the historical water discharge data series (including construction of reservoirs and land use changes).

2.5.3 Database development and conduct of field visits

The same database development strategy as outlined for mainstream floods (Section 2.3.3) applies here. The same holds for the field visits, with the following addition. When lateral floodwater inflow is substantial and can only be estimated from rainfall records, it is important to visit and inspect the rainfall/climatic monitoring stations in and around the contributing area. To judge the suitability of a monitoring station for use in rainfall‐runoff modelling, the visit should include inspections and review of the monitoring station layout, equipment exposure, operating practices and maintenance. 2.5.4 Data validation and processing

Data validation and processing should follow the same steps as presented in Section 2.3.4. If lateral floodwater inflow measurements need to be derived using rainfall‐runoff modelling (SWAT), rainfall and climatic data of the area of the lateral inflows needs to be validated and processed. This involves that the historical series are first screened graphically by comparison with nearby monitoring station records. Erroneous date entries can be identified by tabular comparisons. Monitoring practices during weekends and holidays should be inspected and data should be adjusted where necessary. Next, the data should be compared with data collected by the nearest neighbouring monitoring stations and also subject to double mass tests. Series data should be adjusted when necessary and completed where data is missing, using estimates based on data from reliable measurement stations. The resulting data series can then be converted to areal rainfall for sub‐basins used in the SWAT rainfall‐runoff model. The data series can then be compared with the series used earlier for the modelling. Where there are differences, it may be necessary to recalibrate the SWAT model for the areas where it is to be applied. Note that in several areas, rainfall‐runoff modelling may not be needed to estimate the lateral floodwater inflow, such as in the Xe Bang Fai River example discussed in Section 2.4.7. 2.5.5 Hydrological hazard assessment

The hydrological hazard is determined by peak flow and floodwater volume on the Mekong River. These can be used to estimate the floodwater levels at the confluence, and the tributary flood volumes and peak flow on the tributary, as well as their interrelation. A matrix of flood hydrographs can be plotted to cover the full spectrum of possible hydrographs starting with the peak flow on the Mekong River for floods with return periods of T ≥ 2 years. The development of this matrix requires the following tasks: 1. Creation of homogeneous floodwater discharge series for the key monitoring stations at the

upstream boundaries of the hydraulic model on the tributary and the Mekong River. 2. Estimation of the marginal distributions of annual flood peak water levels and flood

volumes (between June 1st and November 30th) at these upper boundaries (on the Mekong River and tributary). The observed distributions of flood peak water levels and flood volumes are generally well represented by a GEV distribution.



3. Generation of the following regression equations between peak water level flows and flood volumes with their standard errors:

Floodwater volume on the Mekong River (VM) as a function flood peak discharge on the Mekong River (QpM): VM = f1(QpM);

Floodwater volume on the tributary (VT) as a function of floodwater volume of the Mekong River (VM): VT=f2(VM); and

Flood peak discharges on the tributary (QpT) as a function of floodwater volume on the tributary (VT): QpT=f3(VT);

Note that these equations are not required when tributary peak flows do not play an important role. If, on the other hand, the tributary flood volume is not critical, it may be replaced by the tributary flood peak; see Section 2.4.1;

Plotting of the regression equations with year‐numbers (return periods) to the grid points and additions of lines, parallel to the regression line at distances of +/‐ 1x Se and +/‐1.96 x Se.

4. Specification of the matrix of flood hydrographs for distinct flood peak discharge values. This process requires:

Selection of flood peak discharge values and volumes (65=30) and associated hydrographs for the Mekong River in the same manner as in Section 2.4.7;

Determination of range of tributary flood volumes for each of 30 selected Mekong River flood volumes from the scatter plot about the regression line for the tributary flood volume – Mekong River flood volume relationship, generating, using the standard errors: for each selected Mekong River flood volume a low (R‐1.96 SeVT), medium

(designated as Rf2) and a high tributary flood volume (R+1.96 SeVT). This results into 65 x 3 = 90 combinations of Mekong River flood peak, flood volumes and tributary flood volumes (producing a 3D‐grid, as shown on the next page); and

Derivation of 90 synthetic hydrographs, by scaling observed hydrographs in the same manner as in Section 2.4.7.

The steps in the development of the matrix up to a level of 3 variables are depicted graphically in Figure 2‐15. As for mainstream floods, changes in the stage‐discharge ratings (mainstream and/or tributary) due to river bed variations should be taken into account by repeating the model simulations with different values for hydraulic roughness. The roughness change will be commensurate with the maximum observed changes in the water level for bank full discharge relative to the model values used for the calibration.



Figure 2‐15 Development of the flood simulation matrix for combined floods.

2.5.6 Flood hazard assessment using a hydraulic model

The hydrological hazard in terms of statistics and relations between:

flood peak flow and flood volume on the Mekong River, and

flood volume and/or flood peak flow on the tributary is transformed into water levels in the rivers and floodplains in a hydraulic model of the tributary confluence with the Mekong River. The model requirements are specified in Section 2.3.6. Data from the 30 selected Mekong River hydrographs and 90 tributary floods can then be input using the model for the particular river and floodplains system. The water levels that are computed as functions of time and space can be placed in the database. If relevant, in view of river bed changes, the simulations can be repeated at least twice with different specifications for hydraulic roughness and downstream stage‐discharge relations to accommodate the conveyance changes. These should be added to the simulation database before using a Monte Carlo simulation procedure. Simulations are also necessary to analyse any planned construction. The final step is the execution of a Monte Carlo simulation, as described in step 5 of Section 2.3.4. The only difference here is that in this case a 3D‐interpolation is required instead of a 2D interpolation.

T = 2

T = 5

T = 10

T = 25

T = 50

T = 100

Peak flow le

vels

Selection of Flood Volumes

very

low

low

me

diu

m

hig

h

very

hig

h

Step 2

R-1.96 Se R-Se R R+Se R+1.96 Se

R = regression equation V=f(Qp)

Se = standard error

high volume tributary flood

medium volume triburaty flood

low volume tributary flood

Step 3

For each selected main stream flood select 3 tributary floods on volume (high, medium and low)

Step 1

Selection of 6 peak flows from annual maximum discharge distribution GEV

T= 2, 5,10, 25, 50, 100 years

MRC Flood

Guidelines

2.5.7

Combinewater levarea of 1which is and Mukothers atmonitoriamountsBetweenNoy RiveRiver at N

Figure 2‐1

Table 2‐6

Mahaxai

Nakhon P

Mahaxai

Nakhon P

Floods inwater lev

d Management a

Flood Risk Asse

Example: co

ed flooding tavels are affec10,240 km2 alocated betwkhdahan, That Mahaxai anng station sus to 8,560 kmn Mahaxai aners. The monNakhon Phan

6 Hydrolog

Avera

River at N

Phanom

Phanom

n the Xe Banvels on the M

and Mitigation P

essment

ombined floo

akes place incted by backwnd dischargeween the key ailand downsnd Ban Xe Baurveys an arem2. These acd Ban Xe Bannthly averagenom are prese

ical map of Low

age monthly an

Nakhon Phanom

Jan

60.0

6,277

Aug

2,370.7

54,457

ng Fai River Mekong River

Programme Com

ods in the low

the floodplawater from tes its waters Mekong Rivetream. The fang Fai at Higea of 4,520 kmccount, respeng Fai, the Xee flows in theented in Tabl

wer Xe Bang Fai

nd annual flow (

m.

Feb M

41.5 3

4,469 4

Sep

1,687.9 6

51,055 29

may occur fr. The water

mponent 2: Struc

‐ 42 ‐

wer Xe Bang F

ains along thehe Mekong Rinto the Meker monitoringflow in the Xeghway Bridgem2, whereas ectively, fore Bang Fai Rive Xe Bang Fale 2‐6.

i River basin.

(MCM) of the X

Mar Ap

37.2 34

4,130 4,0

Oct No

635.9 240

9,808 14,8

rom June tolevels in the

ctural Measures

Fai River Foca

e lower Xe BRiver. The Xe kong River atg stations: Nae Bang Fai Rie 13. (See Figthe drainage44% and 84ver is joined ai River at M

Xe Bang Fai Rive

pr May

4.7 90.8

000 6,492

ov Dec

0.5 105.8

860 9,111

early OctobXe Bang Fai

and Flood Proof

l Area (Lao P

ang Fai RiverBang Fai Rivt That Phanoakhon Phanoiver is measugure 2‐16.) Te area at Ban 4 % of the eby the Nam

Mahaxai and t

er at Mahaxai a

Jun

565.6

2 17,722

Year

8 7,504.4

1 238,376

ber, coincidinRiver at High

fing

May 2010

PDR)

r, where the ver drains an om, Thailand m upstream ured, among The Mahaxai Xe Bang Fai

entire basin. Oula and Se the Mekong

and Mekong

Jul

1601.4

35,993

ng with high hway Bridge



13 at Ban Xe Bang Fai are seriously affected by backwater from the Mekong River, while Mahaxai is backwater free. Hydrological hazard assessment

The hydrological hazard in the lower Xe Bang Fai River is determined by the floodwater discharge of the river at Mahaxai and the lateral floodwater inflow further downstream as well as by the water levels on the Mekong River at the confluence near That Phanom. The following data is available to describe the hazard:

The Xe Bang Fai River floodwater discharge records for Mahaxai for the period 1988‐2006.

River flow measurements at Ban Xe Bang Fai. The Ban Xe Bang Fai data establishes the relationship between tributary water inflow along the river and at a specific point, given the backwater effects on the stage‐discharge relation at Ban Xe Bang Fai. Tributary water inflow between Mahaxai and Ban Xe Bang Fai can then be estimated as a function of the flow at Mahaxai, without having to rely on questionable inflow estimates from a very limited number of rainfall monitoring stations.

The Mekong River floodwater discharge at Nakhon Phanom and the flow from the Xe Bang Fai River. The water levels at That Phanom can be determined by adding these two data sources and considering the conveyance capacity downstream of the confluence, represented by the stage‐discharge relation of Mukhdahan. The Nakhon Phanom floodwater discharge series data cover the period 1924‐2005.

The discharge series data have been thoroughly validated using similar approaches to those

explained in Section 2.3.4. Where necessary, the data has been adjusted. (See the Xe Bang Fai

River Flood Hazard Assessment (FHA) Report for details.) Using this data, the hydrological hazard in terms of marginal distributions of peak water flows and flood volumes and their interrelations have been quantified. Statistics for Nakhon Phanom are presented in Section 2.4.7. Statistics for Mahaxai are presented in Table 2‐7, Figure 2‐19 and Figure 2‐20. The matrix of flood hydrographs for distinct discharge peak water levels can then be specified, starting with the flood peak water levels at Nakhon Phanom. This is described above in Section

2.4.7. The result consists of 30 (65) combinations of peak water level discharge and flow volumes and 30 associated water discharge hydrographs. For each of these 30 Mekong River flood volumes, 3 Xe Bang Fai River flood volumes can be selected at distances ‐1.96 SeVXBF, 0,+1.96 SeVXBF from the Xe Bang Fai River flood volume‐Mekong River flood volume regression equation using the standard error measure, SeVXBF. (See Figure 2‐20.) The result is a 3‐dimensional grid for peak discharge and volume in the Mekong River (at Nakhon Phanom) and flood volume in the Xe Bang Fai River (at Mahaxai) as depicted in Figure 2‐17. The grid contains

90 points (653), meaning 90 discharge hydrographs for Mahaxai are required for the Monte Carlo analysis. This can be achieved through rescaling of observed hydrographs, as described in Section 2.4.7. Figure 2‐18 shows an example of a scaled hydrograph.



Table 2‐7 GEV‐parameters, flood peak discharges and flood volumes (June‐November) for distinct

return periods in the Xe Bang Fai River at Mahaxai.

Parameter Flood peak

discharge

(m3/s)

Flood Volume

(MCM)

k

α

u

0.341

498

1,614

0.221

2,304

6,105

T (years)

2

5

10

25

50

100

1,757

2,177

2,398

2,626

2,765

2,881

6,916

9,045

10,188

11,386

12,126

12,755

Figure 2‐17 Impression of the 3‐D grid for peak discharge and flood volume in the Mekong River (at Nakhon

Phanom) and flood volume in the Xe Bang Fai River (at Mahaxai). The blue dots only show the

‘outer side’ of the grid; the full grid contains 90 such dots.

28

30

32

34

36

200

250

300

3502

4

6

8

10

12

14

peak discharge Mekong [1000m3/s]

3D grid of volumes and peak discharges

Volume Mekong [BCM]

Vol

ume

Se

Ban

g F

ai [

BC

M]



Figure 2‐18 Scaling of an observed hydrograph for the Xe Bang Fai River at Mahaxai.

14/05 03/06 23/06 13/07 02/08 22/08 11/09 01/10 21/10 10/11 30/110

500

1000

1500

date

disc

harg

e [m

3 /s]

(scaled) hydrograph at station Mahaxai; year: 1988observed peak: 1440 [m3/s]; scaled peak: 1255 [m3/s]

observed volume: 4.0 [BCM]; scaled volume: 3.5 [BCM]

observed

scaled



Figure 2‐19 GEV‐fit to marginal distributions of annual maximum discharge and flood volume at Mahaxai

(Xe Bang Fai River).

GEV-fit to annual maximum discharge in Se Bang Fai at Mahaxai

500

1,000

1,500

2,000

2,500

3,000

3,500

-2.00 -1.00 0.00 1.00 2.00 3.00 4.00 5.00 6.00

EV1 reduced variate

An

nu

al m

axim

um

dis

cha

rge

at M

ah

axa

i (m

3/s

) Annual maximum discharge

GEV-distribution

GEV-fit to annual flood volume in Se Bang Fai at Mahaxai

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

-2.00 -1.00 0.00 1.00 2.00 3.00 4.00 5.00 6.00

EV1 reduced variate

An

nu

al f

loo

d v

olu

me

at

Ma

ha

xai

(M

CM

)

Annual Flood Volume

GEV-distribution



Figure 2‐20 Flood volume – peak discharge relation for Nakhon Phanom and flood volumes relation between

Mahaxai and Nakhon Phanom, without and with Se‐related lines.

Flood hazard assessment

The 90 flood hydrographs of the Mekong River and Xe Bang Fai River selected above can be input to the hydraulic model of the Xe Bang Fai River delta model for conversion of water flow data into water level data. The model layout and boundary conditions are schematically displayed in Figure 2‐21.

Correlation of flood volume Se Bang Fai (Mahaxai) and Mekong (Nakhon Phanom)

Fvol,Mahaxai = 0.0335Fvol,NPhanom + 52.8

R2 = 0.58

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

50,000 100,000 150,000 200,000 250,000 300,000 350,000

Flood volume Nakhon Phanom (MCM)

Flo

od

vo

lum

e M

ahaxa

i (M

CM

)

Correlation of flood volume Se Bang Fai (Mahaxai) and Mekong (Nakhon Phanom)

Fvol,Mahaxai = 0.0335Fvol,NPhanom + 52.8

R2 = 0.58

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

50,000 100,000 150,000 200,000 250,000 300,000 350,000

Flood volume Nakhon Phanom (MCM)

Flo

od

vo

lum

e M

ahaxa

i (M

CM

)

VM+1.96 Se

VM-1.96 Se

VM=f(VNP)



Figure 2‐21 Xe Bang Fai River hydraulic model with boundary conditions.

The morphological development of the river bed, particularly of the Mekong, is taken into consideration by adjusting the hydraulic roughness of the river in accordance with the level changes for bank full discharge relative to the calibrated situation. The equation that can be

used is Manning’s equation, as described in Section 2.4.5 and is slightly rewritten here as: h + ∆h Nnew = ncal (‐‐‐‐‐‐‐‐‐‐‐‐‐)

5/3 (2.8) h where: h = water depth at Qbankfull [measured in meters, m] ∆h = change [m] In addition, the Mukhdahan rating should be adjusted to eliminate unwanted backwater effects on the floodwater level data for That Phanom. The hydraulic model can be run with the same hydraulic boundaries while adding the adjusted Manning‐roughness values. Monte Carlo sampling of the model boundaries interpolation can be run for computed water levels for each grid cell in the hydraulic model to obtain a frequency distribution for the water levels at each cell. Figure 2‐22 shows an example of the frequency distribution at a location in the floodplain of the Xe Bang Fai River, derived using the Monte Carlo random sampling method.

Me

kong

That Phanom

Se Noy

Se Bang Fai

Q

Q

QH

Mahaxai

Mukhdahan

Q81

Q38

Q35

Nakhon Phanom

Q

MRC Flood Ma

Guidelines Flo

Figure 2‐22

Flood hazar The proceddepth and period of T

anagement and

ood Risk Assessm

Derived freq

Xe Bang Fai R

rd mapping

dure for floodextent map = 100 years.

Mitigation Prog

ment

uency curve fo

River.

d hazard mapof the Lowe

gramme Compon

or maximum an

pping is descer Xe Bang Fa

nent 2: Structur

‐ 49 ‐

nual water leve

cribed in Secai River is sh

al Measures and

els at a location

tion 2.7. As hown in Figu

d Flood Proofing

n in the floodpla

an example re 2‐23 , for

g

May 2010

ain of the

the flood r a return

MRC Flood

Guidelines

Figure 2‐2

d Management a

Flood Risk Asse

3 Flood dep

and Mitigation P

essment

pth and extent

Programme Com

map of the Low

mponent 2: Struc

‐ 50 ‐

wer Xe Bang Fa

ctural Measures

i River, for a ret

and Flood Proof

turn period of T

fing

May 2010

T = 100 years.



2.6 Flood Hazard Assessment in the Mekong Delta

2.6.1 General

The floods in the Cambodian floodplains and Mekong Delta (see Figure 2‐24) are classified as a special type in the LMB due to their specific external and internal boundary conditions and the delta’s unique hydraulic infrastructure. Whereas the flood levels near Kratie are mainly dependent on the Mekong River floodwaters and a small amount of backwater from the Tonle Sap Lake, additional factors play a role further downstream. In the Mekong Delta’s downstream areas, flood levels are dependent on upstream and lateral inflows, local rainfall and downstream water levels where the river empties into the sea. Floodwaters in the Mekong Delta are conveyed via the Mekong and Bassac Rivers and their floodplains, including the colmatage canal system, which diverts and regulates the flow of floodwaters from and into the River. In the Mekong Delta the river regime is modified by seasonal water storage in the Tonle Sap Lake and in the Mekong River floodplains, creating slowly rising and falling water levels. Essentially, the hydrological hazard in the delta is the result of upstream and lateral inflows, local rainfall and downstream water levels at sea. The flood hazard in the Mekong Delta is determined by the frequency of flood water levels and flow velocities in the delta. These can be derived from the hydrodynamic delta model based on the ISIS‐software system. This model covers the Mekong Delta downstream of Kratie to the Gulf of Thailand and East Sea. It has been developed to compute water levels, flow velocities and river discharges in the delta as a function of the hydrological hazard components. Assessment of the flood hazard in the Mekong Delta downstream of Kratie benefits from the existence of a relatively long historical water discharge data series for Stung Treng, just upstream of Kratie. Furthermore, for the tributary inflow further downstream and to the Tonle Sap multi‐year data series have been created preserving the serial and cross‐correlation with the Mekong River flow. The Stung Treng time series data, which cover the period 1910‐2006, have been used as boundary conditions for a hydrodynamic model (based on the ISIS‐modelling package) to derive a 97‐year series of water levels in flood‐prone areas. The ISIS model also adds local rainfall, evaporation, water use and tidal conditions at the Gulf of Thailand and the East Sea for the year 2000. Probabilities of flooding and related damages for return periods from 2 to 100 years can be derived directly from the series of water levels and depths computed by the model. The steps required for the flood hazard assessment are similar to those in other hazard assessments and include data collection and conduct of field visits, database development, hydrodynamic model development, and then flood hazard assessment followed by the preparation of flood maps. 2.6.2 Data collection and conduct of field visits

The hydraulic model is based on conditions in the Mekong Delta in the year 2000. To assess the flood hazard for development scenarios relative to the base case, it is necessary that the hydraulic infrastructure pertinent to the scenarios is reflected in the model schematization with different potential constructions and for those that have occurred since 2000. Hence, detailed information is required on layout, dimensions, capacities and operation of the hydraulic infrastructure and flood protection works relative to the model conditions. Field visits can be used to compare the currently existing infrastructure with its schematisation in the model. The model needs to be updated to reflect key changes.

MRC Flood

Guidelines

As for thand dowhydrolog

Figure 2‐2

d Management a

Flood Risk Asse

he hydrologicwnstream bogic conditions

4 Hydraulic

and Mitigation P

essment

cal boundary undary conds in and aroun

c infrastructure

Programme Com

conditions, titions are a nd the Mekon

of the Mekong

mponent 2: Struc

‐ 52 ‐

the 97‐year trealistic repng Delta.

g Delta.

ctural Measures

time series dpresentation

and Flood Proof

data of upstreof the variab

fing

May 2010

ream, lateral bility of the

MRC Flood Ma

Guidelines Flo

Figure 2‐25

2.6.3 D

A databasethe followin

1. A modelayout a

2. A resultand pote

3. A digital4. Data for5. Flood m

anagement and

ood Risk Assessm

The schemat

large numbe

cross‐section

Database deve

has been deng types of in

el database, nd operation database, inential infrastl elevation mr water levelsmaps.

Mitigation Prog

ment

tisation of the M

er of hydraulic n

ns.

elopment

eveloped for ncluded datab

including schns, and hydroncluding the sructure constodel (DEM) os, flood depth

gramme Compon

Mekong Delta u

nodes downstre

flood hazardbases and dat

hematisationological boundsimulated watruction scenof the Mekonh and flood d

nent 2: Structur

‐ 53 ‐

using the hydro

eam of Kratie, n

assessment ta:

n of the hyddary conditioater levels annarios as inpung Delta. uration for d

al Measures and

odynamic delta

next to floodpla

in the Meko

raulic infrastons. nd flow velocts for flood h

istinct floodin

d Flood Proofing

model (ISIS) in

ain ‘reservoirs a

ng Delta that

tructure, its

cities of the bhazard assess

ng return per

g

May 2010

volves a

and river

t includes

structure

base case sments.

riods.



2.6.4 Hydrodynamic model development

The hydrodynamic model of the Mekong Delta is based on the ISIS modelling system for the simulation of unsteady water flows in river channel networks. It provides an implicit numerical solver for the de Saint Venant equations for 1D flow. The model computes water levels and floodwater discharges at select intervals on a non‐staggered grid. Floodplains are included as storage cells connected to the stream channels via two‐way weirs. The system was introduced to the MRC under the Water Utilisation Programme (WUP‐A) and now serves as part of the Decision Support Framework (DSF). The model covers the MRB from Kratie, Cambodia to the East Sea, including the Tonle Sap Lake and its floodplain, the Cambodian floodplains, and the Vietnamese sections of the Mekong Delta. The model used in FMMP‐C2 is based on Delta Model Version‐2009, with adaptations and recalibrations of the original model. The schematisations of the Cambodian and Vietnamese sections of the Mekong Delta that are used in the model are presented in Figure 2‐25. The boundary conditions of the model are based derived from the following data, with each category then described in turn, below:

upstream discharge at Kratie;

tributary inflows to the Tonle Sap and Tonle Sap Lake;

tributary inflows to the Mekong River d/s of Kratie and the Tonle Sap;

rainfall;

evaporation;

water use; and

downstream boundary characteristics at the Gulf of Thailand and East Sea.

Upstream discharge at Kratie

A 97‐year discharge series has been established for Kratie, based on the record of Stung Treng which is available since 1910. For several reasons preference has been given to the series of Stung Treng, including length of record, absence of gauge shifts and backwater effects, stable and predictable hydraulic control for high flows and availability of discharge records. The discharge measurements of Stung Treng, transferred to Kratie reveals excellent agreement with the current meter measurements at the latter location in the past. A serious problem and source of confusion, however, is the difference between stage‐discharge ratings according to conventional discharge measurements and those based on Acoustic Doppler Current Profiler (ADCP) records in use at Kratie since 2002. The ADCP records, due to erroneous referencing of the measurements, lead to lower discharges for the same water levels than the ratings based on conventional discharge measurements, the differences increase for the higher stages. Therefore, unless concurrent current meter and ADCP‐measurements are taken at Kratie and Stung Treng, and discharge rating curves are updated, it is advised using only the Stung Treng records, based on conventional measurements. The Stung Treng records are consistent with measurements of the water flow at Pakse on the Mekong River. Since the hydrodynamic model has been calibrated on recent discharge series at Kratie with flood volumes that are about 8% less than those at Stung Treng, the historical Stung Treng data used in the model can be corrected to reflect these differences, using a reduction of 8%. (Note that this 8% reduction is not applied to the data and statistics presented below for Stung Treng, itself.) The annual maximum discharge series and flood volume of the Mekong River at Stung Treng are shown in Figure 2‐26 and Figure 2‐27.



Figure 2‐26 Annual maximum water discharges at Stung Treng, 1910‐2006.

Figure 2‐27 Annual flood volumes at Stung Treng, period 1910‐2006.

Tributary inflow to the Tonle Sap Lake

The area draining into the Tonle Sap Lake, bounded by Cambodia’s National Highways 5 and 6 that surround the lake, encompasses 68,830 km2 and includes the drainage areas of 13 Stungs (rivers). Water inflow data series of daily discharges are available for the years 1997‐2004. A model has been developed to extend the time series data to 1910‐2006, using contemporary data to determine the relationship between flows at Stung Treng and then applying data from Stung Treng to generate the missing data at other sites. The estimates rely on a multiple

Annual maximum discharge in the Mekong at Stung Treng, period 1910-2006

20,000

30,000

40,000

50,000

60,000

70,000

80,000

90,000

1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Pea

k d

isch

arg

e (m

3 /s)

Annual flood volume in the Mekong at Stung Treng, period 1910-2006

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

400,000

450,000

500,000

1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Flo

od

vo

lum

e (M

CM

)



regression equation for monthly tributary flows. The flows have been regressed on the tributary inflows of the previous of month to reflect the serial correlation of the sites, and on the Mekong River flow at Stung Treng in the same month to preserve the cross‐correlation. Following standard statistical practices, a normally distributed random number was added to preserve the random variance that would normally exist between the two data sets. The monthly flows were then disaggregated to daily flow values based on their degree of agreement with data from the years in which there were existing measurements. These daily flow measurements from years with actual data were then scaled per month to the required generated value. The resulting seasonal inflow to the Tonle Sap Lake is shown in Figure 2‐28. The average monthly inflows as a percentage of the annual totals in comparison to the Mekong River discharges at Stung Treng are presented graphically in Figure 2‐29.

Figure 2‐28 Seasonal hydraulic inflow to the Tonle Sap Lake, estimated from 1910‐2006, with recorded data

for 1997‐2004.

Figure 2‐29 Average monthly flow regime of the Mekong River at Stung Treng and of the Tonle Sap Lake as

percentage of the annual totals.

Seasonal inflow (May to December) to Tonle Sap Lake

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005

Se

aso

nal

in

flo

w (

May

-De

cem

be

r) (

MC

M)

Comparison of runoff regime of Mekong at Stung Treng and Tonle Sap Lake inflow

0

5

10

15

20

25

30

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Mo

nth

ly f

low

vo

lum

e as

per

cen

tag

e o

f an

nu

al t

ota

l (%

)

Mekong at Stung Treng

Tonle Sap Lake inflow



Tributary inflows to the Mekong River

Apart from the inflows to the Tonle Sap Lake, the Delta Model also requires inflow series data for the Prek Te, Prek Chhlong, Prek Thnot, East Vaico River, and West Vaico River tributaries. The daily water flow series for the period 1985‐2006 is available from the DSF files created by the SWAT model. Since the SWAT‐series data show no correlation with the flows in the Mekong River, a pragmatic approach was used to extend the data series; applying a repetition of the time series 1985‐2006 also for the years 1910‐1984 to create a 97‐year series. Rainfall

Daily rainfall series of 9 locations in Cambodia and of 5 locations in Viet Nam are required as inputs to the hydraulic model for the Mekong Delta. Data is available for these locations for 1985‐2006. The annual maximum daily values generally are in the order of 100 to 150 mm, occasionally with larger values up to 400 mm, such as in Can Tho in 1985. Analysis of this data shows weak or no correlation between that seasonal rainfall at the selected locations with water inflows to the Tonle Sap Lake, and no correlation at all with the flows in the Mekong River. As with tributary inflows, a repetition of the series 1985‐2006 was applied to create a data set covering the years 1910‐1984. Evaporation

The model requires evaporation data from the same locations as the rainfall data. For the locations in Viet Nam, a daily data series is available from 1985with the exception of 2002. The data series available for Cambodia are shorter. In view of the limited variability of potential evaporation in a particular month from year to year, long term monthly averages have been applied where daily evaporation data was unavailable. Water use

There are 128 different sites included in the Mekong Delta model where water is used for agriculture, domestic and industrial purposes. The variation in water use varies from about 1,400 m3/s in January down to almost 0 m3/s at the end of September. The total annual water usage amounts to 16.5 Billion Cubic Meters (BCM). During the flood season, the demand for water is about 4.4 BCM in total, i.e., less than 300 m3/s. Sea boundary characteristics

There are 19 different water level boundaries defined in the Delta Model. These boundaries are taken from hourly observations made at the 6 monitoring stations listed in Table 2‐8. The hourly observations used in the Delta Model are records taken from the year 2000. A comparison with historical records demonstrates that the use of this series does not introduce a bias in peak water levels.

Table 2‐8 Overview of water level monitoring stations at coastal boundaries in Viet Nam.

Monitoring

station River Province Remark

Rach Gia

Song Doc

Ganh Hao

Cai Lon

Song Ong Doc

Ganh Hao

Kien Giang

Ca Mau

Ca Mau/Bac Lieu

Waters drain into the Gulf of Thailand

As above

Waters drain into the East Sea

My Thanh Bassac Soc Trang/Tra Vinh Waters drain into the East Sea

Ben Trai

Vam Kinh

Cua Cung Hau

Cua Dai

Tra Vinh/Ben Tre

Ben Tre/Tien Giang

Southern Mekong River outlet, with waters

draining into the East Sea

Northern Mekong River outlet, with waters

draining into the East Sea

MRC Flood

Guidelines

2.6.5

The ISIS regressiomain outmodel arusing thecan thenthe data of water over the

2.7

2.7.1

The MRClevels (+calculateis an imaunobstruthe river river whe(See Figu(DTM) odeterminflooding seasonalagricultudamage

Figure 2‐3

d Management a

Flood Risk Asse

Flood hazar

model is uon equations tputs of the mrea. For eache model outpn be generateseries of ann levels is, by 97 year perio

Flood haza

General

C uses the ISMSL) in partsed for (amongaginary line pucted. A rese(mostly embere water levure 2‐30.) Aftor Digital Elened. (See Figfrequencies maximum lure, housing adata.

0 Water lev

interpolat

and Mitigation P

essment

rd assessmen

used to simuand the use model are esth location, thputs. The ested using the Gnual flood mdefinition, aod of actual d

ard mapping

SIS hydraulic s of the Mekg others) ISISperpendicularervoir node rbanked) with vels are eventer subtractinvation Modegure 2‐31 and (return perlevels up to and infrastruc

vel map, based

tion of ISIS nod

Programme Com

nt

ulate data sof the 97 yeatimated watehe series of 9timates of thGringortens faxima coverspproximatelydata.

model (and kong River Ba river cross‐sr to the riverepresents a an even wat

n, through a png the grounel (DEM) thed Figure 2‐32riods or prothe first ofcture are bas

on

des.

mponent 2: Struc

‐ 58 ‐

series througar water discer levels at al97 annual me probabilitieformula (see s close to 100y the same as

sometimes tasin (see presection nodesr in an area w‘reservoir’ orter level. A GIprocess of innd levels (+Me water dep2.) Flood maobabilities), fAugust. Floosed on these

Figure 2‐31

ctural Measures

ghout the Mharge data sell river sectioaximum wates of exceedequation (2.50 years, the es the maximu

the VRSAP mevious sections and reservowhere the rivr ‘cell’ (polygIS can be useterpolation o

MSL, stored inths, i.e. the ps can also for annual‐mod damage awater depth

Water depth

water level m

(ground) lev

and Flood Proof

Mekong Deltaeries at Stungns and floodper levels canance of the w5) in Section estimated 10um observed

odel) to calcns). The wateoir nodes. A cer can flow mgon), which isd to identify of the node wn a Digital Teflooding levbe produced

maximum levand flood rismaps and on

h map, generate

map subtracting

els.

fing

May 2010

a based on g Treng. The plains in the n be derived water levels 2.3.4). Since 00‐year cycle d water level

culate water er levels are cross‐section more or less s an area of areas of the water levels. errain Model vels, can be d for various vels and for sk maps for n economic/

ed from a

g the DTM

MRC Flood Ma

Guidelines Flo

Figure 2‐32

The map lathem. The MRC projecdatum, they The mappinA: TransferB: CreationC: Creation The three mtasks are scfollows the 2.7.2 T

The transfefollowing st A‐1, A‐2: The

x/y(CS

A‐3, A‐4, A‐ Som

suiretto a

A‐6, A‐7: In A

(bathe

anagement and

ood Risk Assessm

Depiction of

the water lev

yers in a GIS standard datction is UTM y need to be

ng of the hydr of modellingn of spatial dan of flooding

main steps (chematised ae numbering

ransfer of mo

er of modelliteps:

e hydraulic my‐coordinatesSV). ‐5: me of the IStable tools turn periods. a database, l

Access, sepaased on nodee GIS, explain

Mitigation Prog

ment

how water or f

vel (hydraulic s

need all to htum at the M(northern zoconverted to

raulic modellg results (watata; and maps.

(see Sectionsand numberein the flow ch

odelling resul

ing results (w

model, ISIS, cs, and the cal

SIS output cto determineOtherwise, tike Microsoft

rate data set ID’s). Placinged in the foll

gramme Compon

flood depth is o

imulation + GIS

have the samMRC for the wnes). As the do Indian 54.

ling results cater levels) int

s 2.7.2, 2.7.3ed in the flohart.

lts (A)

water levels)

can export thlculated wate

can be procee e.g. the mhe data can bt Corporation

ts can easily g all of the daowing steps.

nent 2: Structur

‐ 59 ‐

obtained by sub

S interpolation)

me datum andwhole MRB idata sets from

an be dividedto a ‘geo‐refe

3 and 2.7.4) w chart in A

into a ‘geo

he (unique) ner level per n

essed with Mmaximum watbe presentedn’s Access.

be combineata in one tab

al Measures and

btracting the gr

.

d projection ts Indian 195m the hydrau

d in three steperenceable’ ta

and correspoppendix 5. T

‐referenceab

ode identifienode, into a

MatLab softwter levels asd on a spread

d into one table reduces t

d Flood Proofing

round level (DT

to be able to54, while the ulic model use

ps: able;

onding data The descriptio

ble’ table inv

er ‘ID’s’ (namcomma‐sepa

ware, which ssociated witdsheet and tr

able by usingthe work to b

g

May 2010

TM) from

o combine standard e another

sets and on below

volves the

mes), their arated file

provides th certain ansferred

g Queries be done in

MRC Flood

Guidelines

Figure 2‐3

2.7.3

ISIS calcuASCII forhave the ISIS has outlines corresposimultan Due to threservoirGIS‐files a GIS is t The metdeal withDAT‐file,

to B‐12; s B‐1: T

uB‐2, B‐3, I

(ptTf

B‐5, B‐6, A

(plr

d Management a

Flood Risk Asse

3 Example o

scenarios

Creation of

ulations are mrmat). The croeir central (po

also a GIS‐c(polygons). Tond with theeously with t

he lack of coor outlines. Focan be used.ime‐consumi

hodology desh the inaccur the methodo

see Appendi

The data stousing the x/yB‐4: ISIS calculate(See Figure 2processing. Ttable containThese two tafocal area in B‐7, B‐8: A spatial (GI(polygons). Hpolygons is nlayer and resremoved, if n

and Mitigation P

essment

of ISIS data imp

s, various return

spatial data

mainly basedoss‐section noint) coordina

component cThe cross‐sece coordinatesthe DAT‐file u

ordination wor the cross‐s Extracting thing (the DAT‐

scribed belowracies in the ology is essen

x 6), depend

ored in the Ay‐coordinates

es with (amon2‐34.) For theTo accommodning the reseables may be Cambodia an

IS) componeHowever, thenot always oservoir layer necessary.

Programme Com

ported into Acc

n periods (prob

(B)

d on node annodes have x/ates in the DA

containing thction and cens in the DATupdate.

ith the DAT‐fsections, eithhe relevant cr‐file for the LM

w uses the ISGIS‐files. If tntially the sa

ing on the inf

Access table .

ng other datae creation ofdate for this,ervoir nodes, split again fond focal area

ent of ISIS coe relationshipne‐to‐one (incan be adde

mponent 2: Struc

‐ 60 ‐

cess, with node

babilities), and d

nd water leve/y‐coordinateAT‐file, but no

he river crosntral reservoiT‐file, but th

file, the GIS‐fher the DAT‐ross‐section MB contains

SIS GIS‐data ahe ISIS GIS‐fime, but som

formation su

can be impo

a) reservoir nf maps, thesethe Access dand anotheror examiningin Viet Nam)

ontains crosp between tn case the Ged to ArcGIS

ctural Measures

information an

different period

el informationes specified.ot their outlin

ss‐sections (r node coordhe GIS files a

files from ISIS‐file or the (ninformation fmore than 17

and not the Dles have beee steps can b

pplied along

orted into Ar

nodes and (rive two types ata‐table canr containing different are.

s‐sections (pthe calculatioIS is not updS. Duplicate l

and Flood Proof

nd water levels

ds of the year.

n in a data‐fiThe reservoirne coordinate

lines) and thdinates in theare not alwa

S need to be unot always ufrom a DAT‐fi70,000 lines).

DAT‐file. It shen synchronisbe skipped (e

with the GIS

rcGIS and geo

ver) cross‐secof nodes nen be split in twthe cross‐seceas inside the

polylines) andon nodes anddated). The cines or polyg

fing

May 2010

for different

ile (DAT‐file, r nodes also es.

he reservoir e GIS should ays updated

used for the pdated) ISIS ile for use in .

hows how to sed with the .g. steps B‐9

files.

oreferenced

ction nodes. ed different wo with one ction nodes. e basin (e.g.,

d reservoirs d the lines/ cross‐section gons can be

MRC Flood Ma

Guidelines Flo

Figure 2‐34

Creating a riv

To create aunder studconversion The cross‐supdated ancross‐sectiosteps need B‐9: In AB‐10: If o

shosho

B‐12, B‐13: The

datB‐15: The

15 ISIS

anagement and

ood Risk Assessm

ISIS schemat

nodes (dots)

(purple polyg

ver cross‐sectio

a data set of dy, the nodeto a ‘raster’ u

section polylnd do not conon nodes for to be conduc

ArcGIS, a buffonly one ISIS ould be copieould be delet

e polylines inta of the crose polylines wper polyline)S cross‐sectio

Mitigation Prog

ment

tization. ISIS cal

). A spatial com

gons).

n point data se

water levels e informationusing both th

ines each rentain the nodwhich ISIS wcted for the c

fer of 150 m cross‐sectioned to the buffed.

nside the (remss‐section nodwith attached ) representinon node. The

gramme Compon

lculates with (a

ponent of ISIS

et for interpolat

or water depn from ISIS he nodes of th

epresent a hde names, onwater levels hcross‐sections

should be drn node falls ifer. If more t

maining) bufde) attached data shouldng the line, sulayer with th

nent 2: Structur

‐ 61 ‐

mong other da

contains cross‐

tion

pths that covneeds to behe reservoirs

horizontal wanly part of thave been cals:

awn around nside a buffethan one nod

fers should g. then be con

uch that all phe buffers can

al Measures and

ta) reservoir no

sections (blue l

vers every loce prepared fand river cro

ater level. If e polylines cculated. In su

the cross‐secer, the (attribe falls inside

get the same

nverted into aoints have thn then be rem

d Flood Proofing

odes (+) and cro

lines) and reser

cation withinfor interpolaoss‐sections.

the GIS‐filescan be associuch case the

ction polylinebute) data of the buffer, t

e data as the

a series of pohe data of thmoved (delete

g

May 2010

oss‐section

rvoirs

n the area ation and

s are not ated with following

es. the node the buffer

buffer (=

oints (e.g. e original ed).



Creating a reservoir point data set for interpolation Like the cross‐sections, reservoirs represent horizontal water levels. Two of the same steps apply here as for the cross‐sections. B‐11, B‐14: The reservoir polygons with only one ISIS reservoir node inside use the data of that

node. If more than one node falls inside, the reservoir polygon should be deleted (in this case, the GIS‐files have apparently not been ‘synchronised’ with the DAT‐file). The resulting layer contains reservoir polygons with the information of one reservoir nodes attached.

There are two options to obtain the required water levels: 1) the reservoir information is merged with the cross‐section information for joint interpolation to obtain the associated water level (in step C‐2), or 2) only the cross‐sections are used for interpolation. The choice depends on the number of available cross‐section points in the area, the flood mitigation scenario considered, the flood type in the area, and the height of the reservoir embankments in the area. Steps B‐16 and B‐17 can be skipped if option 2 is selected. B‐16: Each polygon should be converted into a series of points on the polygon outline (e.g.,

15 points are sufficient). These nodes have the same data as the polygon and also the original ISIS reservoir node.

Merging the cross‐section point data with the reservoir point data for interpolation B‐17: All original cross‐section nodes (B‐3), the derived cross‐section points (from the

polylines; B‐15), all original reservoir nodes (B‐4), and the derived reservoir points (from the polygons; B‐16) should be put into one layer (C‐1). This layer (which may be split into, for example, a Cambodian and a Vietnamese layer) should be the basis for water level point interpolation (See Section 2.7.4).

2.7.4 Creation of flooding maps (C)

Water level (flooding) maps can be created in two steps: 1) the point data, with water levels (B‐15 or B‐17, depending on the chosen option), can be interpolated to create a raster that will cover the area(s) under study, and 2) the reservoir polygons can be converted into a raster and used to replace the raster resulting from step 1. There are many methods to create a continuous surface based on spatial points with values. The type of data, the spatial distribution of the points, the range in the values etc. determine the appropriate method. A few methods have been considered, including: Inverse Distance Weighting (IDW), Spline, Kriging, and Natural Neighbours (NN). Although no in‐depth analysis has been done, the Natural Neighbours (NN) method and the Inverse Distance Weighting (IDW) method appear to be suitable for the LMB. Both the NN and IDW methods allow for the option of including ‘barriers’ or ‘linear discontinuities, such as embankments or canals in the interpolation. Details on the interpolation methods can be found on the Internet.

Flooding maps can then be created by following the steps presented in Appendix 5. C‐1, C‐2, C‐3: The node/point water level values (e.g. year‐maximum water levels with a return

period of 10 years) may be interpolated with the NN method (the raster size is 50 m as

MRC Flood Ma

Guidelines Flo

in tinp

C‐4, C‐5, C‐6 A r

reqC‐7: The

secthor

C‐8, C‐9, C‐ The

delare(50lev

C‐11, C‐12, If w

sho

Figure 2‐35

The previou

2 Using the Ar

anagement and

ood Risk Assessm

the DTM), reputs. 6: reservoir rastquired water e polygon‐bastion nodes, reizontal water10: e raster can tlineation of te under study0 m x 50 m) hvels with a retC‐13: water depth ould be subtr

Example of a

raster.

us steps resul

rcGIS Map Calc

Mitigation Prog

ment

sulting in a r

ter file can blevel values fsed raster (reservoir noder level.

then be ‘cut othe area of iny). The resulthaving a wateturn period o

maps are reracted from t

a raster, with th

lt in a flood (w

ulator with the

gramme Compon

rectangular ra

be created byfrom the polyreservoirs; C‐es; C‐3) to as

out’ more prnterest (or ust is a raster oer level valueof 10 years) ca

equired, thehe water leve

he reservoir ras

water) depth

e ‘con’ statemen

nent 2: Structur

‐ 63 ‐

aster coverin

y converting ygon attribut‐6) can replasure that the

ecisely using sing several mof the area(s)e. The raster an then be cl

DTM with tel raster, and

ster (angular are

h map; an exa

nt (e.g. con(isnu

al Measures and

g the area w

the polygonse table as rasce2 the poine reservoir are

a “mask” (pomasks to dele) under studyvalues (e.g., assified using

the measure classified.

eas) superimpo

ample is depic

ull([reserv]),[xse

d Flood Proofing

with the point

s to rasters, ster cell valuent‐based rasteas do, indee

olygon) with ete various ay with each ryear‐maximg suitable cla

e of ground

osed on the poi

cted in Figure

ec],[reserv]))

g

May 2010

ts used as

using the es. er (cross‐ed, have a

the exact areas that raster cell um water ss limits.

elevation

nt based

e 2‐36.

MRC Flood

Guidelines

Figure 2‐3

2.7.5

Data gena) Thob) Tho

d Management a

Flood Risk Asse

6 Example o

years.

Data source

nerated by those polygons ose polylines t

and Mitigation P

essment

of a flood dept

es

e ISIS model that representhat represen

Programme Com

h map, here in

(with x/y‐coont reservoirs nt river cross

mponent 2: Struc

‐ 64 ‐

the Mekong De

ordinates) inc(i.e., areas w‐sections (i.e

ctural Measures

elta for a flood

cludes: with horizonta., lines with h

and Flood Proof

with a return p

al water levelhorizontal wa

fing

May 2010

period of 25

l); ater level);



c) Model calculation nodes for the reservoirs and cross‐sections. These nodes (from the DAT‐file) may not coincide one‐to‐one with the polygons and polylines mentioned above if the ISIS GIS‐files are not updated. (See Section 2.7.3); and

d) Calculated maximum water levels (referenced to MSL) for various return periods, for various periods of the year (e.g., ‘early floods’), for various areas (‘focal areas’), and/or for various scenarios (flood risk reduction measures).

Data from the MRC/FMMP GIS includes only one category: e) DTM or DEM data. In 2009 two DTM’s were being used at the MRC; b‐dtm50upd and

dem0603. Both are based on interpolation of 10 m contours (on some map sheets, 20 m contours), spot heights and rivers/streams (using topographical maps with ratios of 1:50,000 and 1:100,000, depending on the area). b‐dtm50upd is a 50 x 50 m raster, while dem0603 is a 100 x 100 m raster. Both have significant areas with errors. In b‐dtm50upd, the areas with errors are two large areas near Kampong Chhnang and Kampong Cham. In dem0603 many smaller areas across the LMB have errors. As dem0603 is used in ISIS and its erroneous areas are smaller, the MRC prefers use of this DEM.

CHA

APTEER 3FLOO

3

OD DAAMAGE ASSSESSMENT

MRC Flood Ma

Guidelines Flo

3 F

3.1 In

This chaptedamages aExisting damare presensummarisedpresented. treated in SSection 3.9 The next Cflood hazar

3.2 C

3.2.1 F

The conseqconsequencdue to hurrby floodingbelongings,ecological sHowever, fSection (3.2 An overviewdivided intoeasily assesa market vcasualties, are not tradto be used a

Table 3‐1

Direct

Indirect

anagement and

ood Risk Assessm

LOOD DAM

ntroduction

er focuses ond of benefimage modelsted (Sectiond in Section In order to Section 3.8. F.

hapter (4) ed analyses to

Classification

lood damage

quences of a ces can be obricane Katrina. It encompa, on public systems, induflooding may2.2).

w of differeno tangible anssed physicalvalue exists fhealth effectded in a maras a standard

Classifica

Tangible

Physical dam

‐ Loss of life;

‐ Housing, str

‐ Infrastructu

‐ Agriculture;

‐ Cultural her

‐ Temporary

‐ Cleaning an

‐ Loss of inco

‐ Industrial pr

‐ Others.

Mitigation Prog

ment

MAGE ASSES

on flood damts (Section 3s (Section 3.4n 3.5). Meth3.6. Section develop damFinally, exam

xplains how o estimate the

n of flood dam

es

flood encombserved aftera in 2005. Thesses a wide rinfrastructu

ustrial producy also have b

t types of dand intangiblely. Tangible dfor the damats or damageket and requd for compari

ation of flood d

age to:

ructure and as

re and public

ritage.

relocation;

d sanitation;

me;

roduction loss

gramme Compon

SSMENT

mage assess3.2) and app4) and econoods for the 3.7 describemage model

mples of dama

the results oe risk.

mages and b

mpass multiplr large flood de term “floodrange of harmure, cultural ction and thebenefits and

amages/conse damage, dedamages canaged goods es to ecologicuire different ison. They are

damages.

ssets;

utilities;

ses;

nent 2: Structur

‐ 69 ‐

ment. It outroaches for dmic and finaassessments how the res, data colleage assessme

of damage a

benefits

e types of dadisasters, incld damage“ remful impactssystems, c

competitive this topic is

equences is epending on be directly por services. cal goods andmethods of e indicated o

Intangib

‐ Health

‐ Enviro

‐ Cultur

‐ Societa

‐ Increa

al Measures and

tlines the cldamage assencial issues inof other ty

esults of damction is alsoents for the

ssessment c

amage. All thluding the flofers to all varon humans, ultural heritstrength of tpresented s

presented inwhether or priced in monOther types d to all kind assessment in the table a

ble

h effects;

nmental dam

al heritage va

al and cultura

sed vulnerabi

d Flood Proofing

lassifications essment (Secn damage asypes of dammage estimateo critical. ThiLMB are pre

an be comb

hese differentooding of Newrieties of hartheir health tage, social the affected eseparately in

n Table 3‐1. Dnot the lossenetary termsof damagesof goods andif monetary ts ‘intangibles

mages;

aluation.

al disruption;

ility of survivo

g

May 2010

of flood ction 3.3). ssessment mages are es can be s topic is sented in

ined with

t types of w Orleans m caused and their systems, economy. the next

Damage is es can be , because s, such as d services terms are s’.

ors.



Another distinction can made between the direct damage, caused by physical contact with floodwaters, and damage indirectly following from a flood. Direct damage includes, for example, damage to buildings, economic assets, loss of standing crops and livestock in agriculture, loss of human life, immediate health impacts, and loss of ecological goods. Indirect damages are damages caused by disruption of physical and economic linkages of the social and cultural systems and the economy3, and the extra costs of emergency and other actions taken to prevent flood damage and other losses (see, e.g., Parker et al., 1987). This category includes, for example, the loss of production of companies affected by the flooding, induced production losses of their suppliers and customers, the costs of traffic disruption or the costs of emergency services. The main damage categories that have been used in the past by neo‐classical economists are: 1. Direct and tangible damages

Damages to properties and infrastructure In the targeted areas, property and infrastructure damage can include damage to houses, schools, offices, commercial and industrial buildings and installations, hydraulic works, energy, transportation and other public infrastructure, and supplies/assets of individuals and businesses. The extent of these damages is usually proportional to the magnitude and duration of the flood. Damages to property are usually calculated as the costs of replacement or restoration of an affected structure and inventories.

Agricultural damages Crop damages depend on the depth, timing and duration of a flood. The bigger the flood, the larger the submerged cultivated area that sustains losses. However, if floods occur after harvesting, there may be no crop damage at all. Damage to livestock and agricultural equipment also depends on the magnitude and duration of the flood. These damages are measured by the costs of replacing livestock or the loss of income due to a distress sale. The damages to agricultural equipment/tools are the costs of replacement or restoration of these assets.

2. Indirect and tangible damages

Costs of temporary relocation and rescue The costs of relocation and rescue are incurred by local government agencies and non‐governmental organisations (NGOs) before, during and after floods to support affected people living during the flood and recovering after the flood.

Costs of cleaning and sanitation Cleaning and sanitation expenses incurred after a flood are mainly labour costs borne by individuals, households, enterprises and institutions to clean their houses and buildings.

Income losses and higher costs of living Losses of income due to disruption of economic activities and/or services by floods effect: (i) individuals, landless labourers, families and enterprises; and (ii) large commercial and industry enterprises that have been partly or fully closed for some period of time during high flood water levels. Higher costs of living due to temporary relocation, the purchase of food and other items that otherwise would be available from a family farm of household,

3 Alternative definitions of indirect damage exist. For example, Merz et al. define indirect damage as damage that occurs outside the flooded area (Merz et al., 2004). E.g. companies can lose supply and demand from the flooded area.

MRC Flood Ma

Guidelines Flo

lack of and dai

Costs oImportaeffects.(increaand livborne dand adwell asbriefly other c

Other care beinternaprotectenvironincludegovernfor peoeconom

3.2.2 F

Flooding in the damageimportant acontrol meareas:

FisheryFlood wan impLow flotiming and nat

AgriculFlood wfloodplfrom cwaste dcan be

CoastaFlood wland ar

BiodiveFloodinwetlandmoneta

anagement and

ood Risk Assessm

basic supplieily activities a

of illness ant indirect . It is noted tsed) cost of testock are rediseases. Theditional expe for sanitatiodiscussed in costs.

costs: Other coming increational laws tions during nmental lossee loss in sociment, negatiople or entemic linkages. T

lood benefits

the MRB caues, it is imporand beneficiaasures need

y sector water provideportant resouood levels reand durationtural fish stoc

tural producwaters have ain especiallyultivated soidecompositiomeasured by

l land expanswater brings ea into the se

ersity conservng plays an imds in the LMary values cal

Mitigation Prog

ment

es as safe waall fall into th

damages cathat part of ttreatment anelated to woese costs inclenditures by on control duSection 3.6 a

categories aeasingly impand treaties disasters li

es, e.g., due al stability, pive impacts oerprises outsThese are om

s

uses not onlyrtant to take al role in maito take thes

es habitat fource for resideduce naturan of floodingcks.

tion a number oy boosting thl; (iii) kill theon; (v) improy comparing h

sion sediments tea. This can e

vation mportant rolMD. There alculated in th

gramme Compon

ater and elecis category.

ategories arethese healthnd medical caorsened enviude additionlocal governmuring a flood.as an example

are elaborateportant in as and particuke flooding.to pollutionproblems in of flooding oside the targmitted from t

y damage, bu these into aintaining natse into accou

r fish breedindents of the al fish stocksg, there is ge

of agriculturhe fertility ofe rats and otve soil structharvests in ye

to river mouteasily be mea

le in maintaire a numbehese studies c

nent 2: Structur

‐ 71 ‐

ctricity, and a

e the loss of effects coulare. Health coronmental cal healthcarement for dise. Methods foe of measure

ed by specialssuring that ularly for sus Other imp, and societaincome distr

on tourism reget areas, ahe study.

ut also benefccount in daural fisheriesunt. Benefits

ng, nursing aLMB; particu. Though beenerally a dir

ral benefits.f acid sulphather pests; (ture; and (vi)ears of differe

ths in Viet Nasured and va

ning biodiver of studies can be applie

al Measures and

additional co

f life and injd be monetizosts due to flonditions ane costs for famease preventor the estimament of one

ists in other the measurstainable devortant indireal disruption ribution and evenue, loss ond other fo

its. Since themage assessm, wetlands, acan be meas

nd growth. Nularly in Camnefits depenect relations

They: (i) brite soils; (ii) fiv) speed up promote soent flood leve

Nam, slowly ealued.

rsity in the Mon these bed.

d Flood Proofing

osts for trans

juries and thzed by estimlooding affecnd increases milies and intion and treation of loss oof these cate

fields. Someres meet covelopment aect damagesin general. Tloss of confof income ororward and

ese benefits cment. Floodiand soil fertilsured in the

Natural fish smbodia and Vnd on the mship between

ings sedimenflush toxic sup the processoil fertility. Thels.

expanding th

MRB, especiaeneficial imp

g

May 2010

sportation

he health mating the t humans in water‐dividuals, tment, as of life are egories of

e of these ncerns in nd rights s include The latter fidence in r revenue backward

can offset ng has an ity. Flood following

stocks are Viet Nam. agnitude, n flooding

nt to the ubstances s of plant his impact

he Delta’s

ally in the pacts, the

MRC Flood

Guidelines

In generrelatively(higher r

3.3

3.3.1

Before stthe damac) the avfactors a a) The g

The interindivtimethe studdeta

b) The The the sconshandinvesone to be

c) The Budga smmea

d) The The alreaapprof dapprexistof a

Selection A key quit would damage come to importancostly, a

d Management a

Flood Risk Asse

al these beny frequently eturn period

General ap

Scope of th

tarting a damage evaluatiovailability of re outlined b

geographic asize of the

rnational scalvidual structue and money)most preciseies with a reiled methods

objectives method and study. If the gsequently lesd single floodstment), morshould consie made.

availability oget or time remall budget asures are nee

availability oeffort or coady exists. Tropriate methdamage funcroaches on ating and availmethod.

n of damage

uestion in dambe desirableassessment, an agreement, seek consnd to focus

and Mitigation P

essment

efits are ass(e.g., annuas) generally b

pproach to e

e analysis an

mage assessmon. This deperesources fobelow.

area e area undele to local areure level. Due) than less dee methods aesearch area s.

the accuracygoal of the sts costly and d protection re precise meider whether

of resources estrictions usand/or little teded.

of data sts of damagThe availabilihod. If, for exctions alreada regional orlable in a cou

categories

mage evaluate to take all this could leant accepted sensus on mon them to

Programme Com

ociated withal or seasonbring fewer b

conomic dam

d level of det

ent it is impoends on a) thr conducting

er investigatieas, hence rae to the fact tetailed approre often restof regional o

y of the resultudy is stratetime‐consummeasures h

easures are nr such detail

ually limit thtime are ava

ge evaluationty of pre‐exxample, adeqdy exists, thr even nationuntry or regio

tion is which of the typesad to extensiby all stakeheasures of eo keep the a

mponent 2: Struc

‐ 72 ‐

relatively smnal floods). Ebenefits.

mage assess

tail

ortant to deche geographicg the study, a

ion can varyanging from bthat very precoaches and thtricted to smor even natio

lts of a specifegy developmming) methodhave to be aneeded. Befoinformation

e precision thilable, quick

n can be sigxisting data quate land‐ushis may facnal scale. Fuon can also ha

categories os of damagesive and costlyholders on whestimation if analysis man

ctural Measures

mall floods aExtreme eve

sment

ide on the scc area of the and d) the av

y from areabasin, countrycise methodshat the resoumall areas unonal size mo

fic study musment and pland might be sussessed, (e.gore investing really effects

hat can be acand easy me

gnificantly reshould detese data or a tcilitate the urthermore, tave a significa

of damages shs listed in taby studies. It ishich damage precise meanageable. It

and Flood Proof

nd processesnts that are

ope and levestudy, b) thevailability of

as of nationy, district, vills require morurces are usunder investigstly have to

st match the nning, a less pufficient. If, og., for a prioin a more prs the results

chieved in a sethods and a

duced if necrmine the chhoroughly deapplication the type of dant impact on

hould be inclble 3.1 into as often morecategories asures are unis important

fing

May 2010

s that occur e infrequent

el of detail of e objectives, data. These

nal or even lage or even re effort (i.e. ually limited, gation, while rely on less

objective of precise (and on the other oritisation of recise study, or decisions

study. If only approximate

cessary data hoice of an eveloped set of detailed data is pre‐n the choice

uded. While account in a e practical to are the most navailable or t that these

MRC Flood Ma

Guidelines Flo

selected castandardisedamage catWithin eacrecreationa Less importtypical daminappropriathe numbeestimating 3.3.2 D

Methods foand housesis widespre1977; Dutta Flood damaa) intensb) numbec) suscep

equati These facto

Figure 3‐1

anagement and

ood Risk Assessm

ategories coned damage ategories are h of the maal use within t

tant categoriemage multiplate to standaer of lives loloss of life).

Direct damage

or the estimas are well estead. (See bela et al., 2003;

age estimatesity of floodiner and type optibility of laions).

ors are presen

General app

depths used

Mitigation Prog

ment

nstitute the ccounting mprioritised: laain categoriethe main land

es can also beier to accourdise in termost. (See, fo

e assessment

ation of direcablished, andlow for furth; Kok et al., 2

s require infog; of land uses aand uses to

nted schemat

roach to damag

to calculate da

gramme Compon

largest partmethod that iand use, infras are sub‐cad‐use catego

e included bynt for them.ms of market or example S

t

ct economic d the use of her explanati005; and Jon

ormation on t

ffected by floflooding (re

tically in Figu

ge estimation, w

amages.

nent 2: Structur

‐ 73 ‐

t of the tots used in thastructure, hategories (e.gry).

y using estim. For “intangvalues, otheSection 3.6

damage to p’damage funion and also kman et al., 2

the:

oods, and theepresented

re 3‐1 and de

with the relatio

al Measures and

tal damage. e Netherlandouseholds, cog., agricultura

ation measurgible” damagr measures cfor a discus

physical objecctions’ to estPenning‐Row2008 for add

eir value; andby floodwat

etailed below

onship between

d Flood Proofing

For examplds the followompanies anal use, urba

res, such as age categoriescan be chosession of met

cts, such as stimate these wsell and Chitional backg

d ter damage

w.

n land use and f

g

May 2010

e, in the wing main nd ‘other’. n use, or

applying a s that are n such as thods for

structures damages

hatterton, round.)

function/

flood

MRC Flood

Guidelines

a D

iifsi(

b D

dnfoi

c S S

fcap

It is possthe totalaccustomall of theintensity

whe

For eachthe dama 3.3.3

Damage function on the 19factor of

d Management a

Flood Risk Asse

Intensity of fData on the inundation cinundation afactors such significant inintensity of f(Section 2).

Number andDifferent typdamage valunumber and flood is a parof primary dinformation o

SusceptibilitySusceptibilityfunctions uscharacteristicadded as varpresented in

ible to descri damage in amed to mathee specific poiny and characte

re: Dmax,

i = r = m = n = hr =

αi(hr ni,r =

damage catage assessme

Damage fun

functions reis given in Fig953 flood (Sef 1.0 in the c

m

i

n

r

D

and Mitigation P

essment

flooding intensity of haracteristicsare the mostas the duratfluence on dflooding and

type of landpes of land usuation. For etype as well rt of damage data from fieon data colle

y of land usey relates ecosing flood dcs, such as floiables in the Section 3.3.

ibe how thesa flooded areaematical symnts in the floeristics of the

x,i = maximumdamage location number number hydraulicparticula

r) = stage dadamage particula

= number

egory a specent (e.g., corr

nctions

late the levegure 3‐2 for dee Kok et al. (hart) when w

iri Dh max,)(

Programme Com

a flood, (i.es) are essentt important ftion, time ofdamage depeits measurem

d uses affecteses that correexample, foras the elevatestimation. S

eld surveys oction and lan

s to floodingonomic damadepth measuow velocity omodelling eq

e elements ina in an equatmbols. The tood area and e flood:

m damage amor land use cin flooded arof damage caof locations ic/flood charar location mage functiofor category ar location r (0of objects of

ific stage damrelating histo

el of damage damage to ho2005).) The fwater depth

rin ,

mponent 2: Struc

‐ 74 ‐

e., the affectetial to damagfactors in mef occurrence,ending on thment is prese

ed land uses,espond with property dation of propeSuch informaor secondary nd use survey

g age to floodures are ther flood duratquations. Fur

n the damagetion. It is simpotal damages the damage

mount for an category rea ategories in flooded areracteristics (

on that exprei as a functio0 ≤ αi(hr) ≤ 1)damage cate

mage functioorical damage

to flooding omes and thefigure shows exceeds 4.5

ctural Measures

ed area, the ge estimateseasuring inte, and water e types of floented in the f

and their vadamage cate

amage, informerties potentiation is normdata, such a

ys is presente

d conditions. e standard tion could alsrther details o

e model can bple to describare a summin all of the c

object or lan

ea e.g., intensi

sses the fracton of flood ch) egory i at loca

on can be dere data with flo

conditions. Aeir contents ithat flood dameters. Ther

and Flood Proof

flood water. The area anensity, but nvelocity can ood. Informaflood hazard

lue egories are amation on thally affected ally availableas public mad in Section 3

Often, (stagtools, but oo be relevanton damage fu

be combinedbe in words foation of the categories re

(3.1)

nd use catego

ty) of the

tion of maximaracteristics

ation r

rived and theoodwater dep

An example on the Netheramage is totare is a large v

fing

May 2010

r levels, and nd depth of non‐intensity also have a ation on the assessment

a key part of he location, by a certain

e in the form aps. General 3.8.

ge) damage other flood t and can be unctions are

d to estimate or those not damages at

elated to the

ory i

flood at a

mum at a

en applied in pth).

of a damage rlands based al (100% or a variability in



damage functions, often different by country. For example, in the United States the Federal Emergency Management Agency (FEMA) has developed the HAZUS risk assessment software model that includes damage functions for different damage categories4.

Figure 3‐2 Example of a stage damage function for houses in the Netherlands (Kok et al., 2005). Note that

the curves for building damage and contents show these as percentages of the total damage.

Influence of flood characteristics on the damage function Most stage damage functions include water depth as the main determinant of direct damage. However, other flood characteristics, such as flow velocity or flood duration can also influence damages. Thieken et al. (2005) investigated the influence of these and other factors, including flood duration, contamination and flood damage preparations, based on data for the 2002 floods in Germany. Other factors, including psychological ones like risk perception of residents in flood prone areas and their preparedness can also influence susceptibility to risk and, hence, damage. Such factors have yet to be included in damage functions − at least explicitly. In the application of damage functions to estimate the damage for agricultural land use (e.g., for crops or rice paddies), other factors besides flood depth are important in assessing damage. Duration of flooding, for example, can extend damage and/or prevent harvesting. There is also a seasonal effect, since damage depends on the stage at which floods occur in the growth and harvest cycle. Damage functions need to account for these damage processes (moment during the year, duration) and characteristics of the land use (such as crop characteristics). Relationship between structural damage and economic damage An important source of economic damage is the structural damage to buildings, especially homes. Figure 3‐3 shows how economic damage is connected to structural damage. The level of structural damage caused by a flood is determined by the physical forces (or loads) and the building resistance (or strength) (Kelman and Spence 2004). The degree of structural damage depends on the intensity and magnitude of the flood (i.e., hydrostatic and hydrodynamic forces), and on the building’s resistance to flooding. To convert the structural damage to economic estimates requires insight in the building’s pre‐disaster market value and the replacement cost. Apart from damage to the structure, the damage estimate in monetary terms should also include damage to the inventory of the structure (for instance items in the building).

4 see http://www.fema.gov/plan/prevent/hazus/ for more details

0,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0

1,1

0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 5 5,5 6

Flood depth (m)

dam

age

fact

or

total damage

damage building

damage content

MRC Flood

Guidelines

In the dedirectly rthe degrdegree obuildingsFor strucdamage during flo

Figure 3‐3

3.3.4

There ar‘absolutefunctionsthe infor 1. The

In thflooddaminfor(agri

2. The

In thfloodpercrisk comrathe

Figure 3‐also showthat in th(or percedamage

d Management a

Flood Risk Asse

escription of related to floree of structof structurals. In Appendixctural damageis also imporood events.

Schema o

Relative and

re two basic e damage’ fus (percentagermation sourc

use of absoluhe absolute dd depth or fage during hrmation, can culture, hous

use of relativhe relative dad depth andentage) of this calculatedpared to theer than in mo

‐4 depicts hows the differehe relative daentage) as avalue follows

and Mitigation P

essment

damage funod charactertural damage damage cax 2 existing me both the dertant for hum

of structural flo

d absolute da

approachesunctions (a se loss of the tces on land u

ute damage famage approflood water historical floobe used to sing, infrastru

ve damage fuamage approd flood damhe maximum by combinie maximum onetary term

ow the differeent formats oamage approa function ofs directly from

Programme Com

ctions, aboveristics. Howeve and the nuan be measumethods for tepth and flowman safety an

od damage ana

amage assess

to damage specific finantotal potentiase, values of

functions foroach, the damlevels and tods combinegenerate equcture, etc.).

unctions for dach, the dam

mage for eac potential dang land use potential das.

ent elementsof the damagach, the damf water deptm the damag

mponent 2: Struc

‐ 76 ‐

e, the econover, it is someumber of daured in the the analysis ow velocity wilnd loss of lif

alysis.

sments

assessment ncial measural loss). The aassets and d

r damage assmage functiothe direct med with dataquations for

damage assemage functionch specific laamage (100%information

amage to ex

s are combinege functions fmage functionh, whereas e function.

ctural Measures

mic damage etimes imporamaged or dnumber of

of structural dl be importane as building

using damare) and one approaches damage funct

essment ns describe t

monetary dam on inundatdamages in

essment n describes tand use cate% losses). Theand asset vpress the da

ed in the twofor both casen produces a in the absolu

and Flood Proof

estimate is drtant to havedestroyed budamaged odamage are snt. The level ogs are import

ge functionsusing ‘relatidiffer in the wions are com

the relationshmage value. ion depth anspecific asse

he relationshegory as a e total value alues. Then, amages as a

o approacheses. The main drelative damute damage

fing

May 2010

described as e insight into uildings. The r destroyed summarised. of structural tant shelters

s; one using ive damage’ way in which mbined:

hip between Surveys on nd land use et categories

hip between fraction (or of assets at this sum is percentage

s. The figure difference is

mage fraction approach a

MRC Flood Ma

Guidelines Flo

Absolute approach

Relative approach

Figure 3‐4

The two apadaptability Absolute danew regionRelative daland‐use caapplied to relative damsusceptibilitdamage apfactors. Absolute ddamage funavailable, th Note that fdistort the district levelikely. 3.3.5 A

Indirect daeconomic liprevent floflooding, indisruption o

Hcha

Hchar

anagement and

ood Risk Assessm

Absolute and

pproaches hay to local data

amage functn, data is reqmage functioategories. If a wider regmage functioties are similpproach is th

amage functnctions alreahe relative da

for both appanalysis and el and the res

Assessment of

mages includinkages of theood damage nduced prodor the costs o

Ad

Susceptibiassets

Hydraulic racteristics

Rdam

Su

Hydraulic racteristics

Mitigation Prog

ment

Approach

d relative dama

ave advantaga.

ions are specquired to derons are morethey have bion for whicons can only ar (e.g., floodhus more sui

tions are alsady exists. If amage appro

proaches, thehide errors. sults are pres

f indirect dam

de such catee economy, aand other l

duction losseof emergency

Absolute depth damage function

lity of s

Value o

Damage

elative depth mage function

usceptibility of assets

V

Damage

gramme Compon

age approaches

ges and disad

cific to the arive the spece generally apbeen derivedch no historibe extendedds to similar itable when

so advantageon the othe

oach may be m

e varying leveIf data at thsented at the

mages

egories as daand the extralosses. The es of their sy services are

Land use information

of assets

Total value of assets at risk

Value of assets

nent 2: Structur

‐ 77 ‐

s.

dvantages in

areas wherecific absolutepplicable sincfor a certaical flood dad to other aragricultural llarge region

eous wherer hand, detamore conven

els of detaile commune e commune le

amages causa costs of emeloss of prodsuppliers and also included

Land use information

al Measures and

Funct

applicability

they have bee damage fuce they look ain region andmage data iseas where dands and strunal difference

a compreheiled data on ient.

at different level is combevel, this kind

sed by disrupergency and ouction of cod customers,d in indirect c

Damage ($)

Damage (%)

100%

d Flood Proofing

tion

to other are

een derived. nctions for tat damage imd event, thes available. damage proceuctures). Thees exist in t

nsive set of asset values

geographic lbined with dad of ‘false’ ac

ption of phyother actionsompanies aff, the costs costs.

Water delevel (m)

Water de

g

May 2010

eas and in

For each that area. mpacts on ey can be However, esses and e absolute he above

absolute s is easily

evels can ata at the ccuracy is

ysical and s taken to fected by of traffic

epth or )

epth (m)

MRC Flood

Guidelines

The lossevery signor road nimportanusing a mrelies onapproachdevelope Today, incertain ineconomi Surveys cfor indireand indifunction indirect e

3.4

For consassessmegovernmthat can risk and c Examplesstandardapproachthat the Based ondevelopetested in An autom

A larproc

The

Ther(e.g.

Diffethe l

d Management a

Flood Risk Asse

es due to busnificant, espenetwork is afnt companiesmodel that inn “input‐outphes are beined for practic

n practice, lonterruption pc activity (ad

can also be uect damagesrect damageof the directeconomic da

Damage m

sistent decisient on a wid

ments in sevebe used acrocost‐benefit c

s are the Hdised damaghes, the landdamage can n the approaed under then Thailand – s

mated approa

rge number ocessed;

underlying da

re may be a , situations w

erent areas coand uses and

and Mitigation P

essment

siness interruecially if impoffected by flos is affectedncludes the eput” models.ng researchecal use.

osses due to period (e.g. aded value) an

useful to det. Eventually, es. With sucht damages. Fmage to the

models

ion‐making ite level, such eral countriesoss countriescalculation.

HAZUS methe model ded use data anbe easily estach developee name of DAee Textbox 1

ach is specific

of damage ca

ata on land u

large numbewith various a

onsidered in d housing typ

Programme Com

ption and indortant econoooding and ha, these losseexisting netw (See, for exed and publ

business intea number of nd the numb

ermine the ithis offers d

h a ratio, it iurther detailLMB are foun

t is desirableas for a rives have deves for the esti

hodology, deeveloped in nd damage futimated if infed in the NeACA (Damage1.

cally attractiv

ategories and

use and value

er of calculatalternative pr

the model upes are compa

mponent 2: Struc

‐ 78 ‐

duced losses mic hotspotsas to be closes can be suwork of linkagxample, Van ished in aca

erruption aremonths) ander of affected

ndirect damadata that cans possible tols on the appnd in Volume

e to have a er basin or at loped standamation of da

eveloped in the Netherunctions areformation foretherlands, ae and Casua

ve, when:

d/or a large

es of assets is

tions and/or rotection sche

use the same arable).

ctural Measures

in other parts are floodeded for a longubstantial. Thges within ander Veen e

ademic litera

e generally ed a cost factd employees.

age categorie demonstrato estimate thplication of the 2C.

standardisethe countryardised methamage due to

the USA (lands (Kok included in ar a certain flo more generlties Assessm

amount of la

available in G

different scemes and me

equation for

and Flood Proof

ts of the econ. If an imporg period or if his effect cann economy. Tt al. (2003))ature but ha

estimated by or that inclu.

es and the dae a ratio bethe indirect dhis approach

d approach level. With thods and sofo flooding as

FEMA, 2003et al., 2005a computer mood scenario ric model/tooment) and ha

and use data

GIS format;

enarios beingeasures).

r damage est

fing

May 2010

nomy can be rtant airport f a cluster of n estimated The method ). Yet, these ave not yet

assuming a des the lost

amage costs tween direct damage as a to estimate

for damage this in mind, ftware tools s well as the

3), and the 5). In these model, such is available. ol has been as also been

a need to be

g compared

timates (i.e.,

MRC Flood Ma

Guidelines Flo

Textbox DACA (Dassessmof land uscenarioways thland usgovernmon a sta The Thaprovince Source:

Leenders

3.5 E

Within the values are

below. Furt

for example 3.5.1 F

There are dfinancial ansingle persoindividual oor regional perspectivepublic polic 3.5.2 E

For damageuses that co The basis founit. The ethat is rela“distort” m

5 Since these what they calshadow pricesubsidies (‘tramarket pricesassume an ‘e

anagement and

ood Risk Assessm

x 1 DACA (D

Damage and Cment model ‘Huses and theos. The tool cat tie damagse planning, mental agencndard approa

i Departmene of Chiang R

“Developmen

s et al., 2009)

conomic and

context of dimportant. S

ther discussio

e, Messner et

inancial vs. e

different pernd economic von or firm, neor firm. Econowelfare, incl

e is the approcy.

stimating val

e assessmentould be poten

or valuation fconomic evaatively easy arket prices a

ideal conditionll the ideal ‘shaes by trying to aansfer payments or influences quality’ of all p

Mitigation Prog

ment

amage and C

Casualties AsHIS‐SSM’. Their economic calculates dirges to their oflood manaies and depaach and trans

t of Water Reai, which incl

nt and Use/D

.

d financial is

damage asseSome of the

on is included

t al., 2007).

conomic valu

rspectives onvaluation. Fineglecting pubomic evaluatiuding impactopriate one t

ues of assets

ts it is imporntially damag

favoured by nluation of goto accomplisand need to b

ns are never fulfdow price’ of thccount for somts’ to particularof major buyerotential buyers

gramme Compon

Casualties As

ssessment), ise assessmentvalue, as werect damage, origins and tagement andartments towsparency in m

esources hasludes several

Demonstratio

ssues in dam

essment the e main (and,

d in Appendix

uation

n damage apnancial evaluablic affairs anions are broats on intangito apply if ca

s

rtant to haveged. The ques

neo‐classical oods and sersh. Howeverbe taken into

filled in a real‐whe good under me of the distorr individuals or rs and sellers ws and sellers.

nent 2: Structur

‐ 79 ‐

ssessment) an

s based on tht takes into all as the diffeproduction lhat are desigd mitigationwards more cmethods and

tested DACA rivers.

n of DACA i

mage assessm

economic vayet, fully un

x 3 and provi

ppraisal, startations considd focusing onder. They seeble goods analculations of

e insight in tstion is what

economists irvices which r, it is widelo account in d

world market, nexamination. Ttions that are egroups), or by ith other agend

al Measures and

nd its applica

he Dutch damaccount the serent charactlosses and ingned to aid d planning Dollaboration data.

A in a pilot ar

n de LMB” (

ment

aluation of dnresolved) is

ded in backg

ting with theder damage frn the actual fek to assess tnd services. Tf flood dama

he values of economic va

s the marketare traded iny recogniseddamage estim

neoclassical ecoThey translate measy to measuretrying to accoudas into ‘compe

d Flood Proofing

ation to Thail

mage and casuspatial distribteristics of flondirect damagdecision makDACA also pdue to its re

rea in Thailan

(Anonymous,

amages and ssues are sum

ground literat

e distinction rom a perspefinancial burdthe impact onThis broader eage are to be

f the objects alue to assign

t price for an n the marketd that severmation5.

onomists try to market prices ine such as taxesunt for monopoetitive’ market

g

May 2010

land

ualties bution ooding ges, in kers in pushes eliance

nd; the

2009;

land use mmarised

ture; (see,

between ective of a den on an n national economic e used for

and land to them.

object or t is a task al effects

calculate nto and listic prices that

MRC Flood

Guidelines

The concwhen it ibecause diminutioduring a (Messne Similarly,depreciareligious a peoplediversity applied. 3.5.3

In an ecobuilding dependinor later, Damage a certainimportaneconomior decrecan also

3.6

3.6.1

Apart froterms,, tcalculatioexample Note thaby those importan Note thamonetarsocietieslife. Valuvisible wthese dislife separ

6 One excedamage as

d Management a

Flood Risk Asse

cept of depres damaged breplacemenon in value oflood are usr et al., 2007

, for a smtion are inaor heritage s (such as ancthat may be

Time value

onomic analythat is wortng on expectand on other

values that an year in thent to take inc growth incrase with chachange.

Loss of life

Introductio

om the estimthere are sevon of damage. In this Secti

at while these currently emnce.

at when damary value or da, different duing the losswhen economscussions, thorately and ind

eption is the apssessment in th

and Mitigation P

essment

eciation shoulby a flood. Rent usually inof somethingsually substit).

all group opplicable anstructures wicient trees an key to surviv

of money

sis it is importh USD 50,0ations peopler factors.

are used in a e past. Whento account rease values,nging condit

and health

n

mation of ecveral approaces in differenon some exis

e approachesmpowered to

ages like the amage sum pecision‐makes of life direcmic ‘market’ ose currentlydependently

pproach to estimhe Netherlands

Programme Com

ld be appliedplacement covolves improg over its prtuted by new

of irreplacead market vath emotionalnd spirit shrinval of a large

rtant to consi000 this yeare have about

certain daman one is estchanges in but values cions (like cha

effects

conomic damches to the et measures. Lsting approac

s and methodconduct dam

loss of life arper fatality), ters and diffectly raises etvaluations ay conducting from the eco

mating loss of li.

mponent 2: Struc

‐ 80 ‐

in order to rosts, alone, covements wroject usable w, more prod

ble assets, alues may nl value or spines in the MRr system. Oth

ider the valur could havet the future a

age model artimating damthe values an also decreanging popul

mages, or coestimation ofLoss of life anches and liter

ds are not yemage assessm

re expressed there is no uerent disciplinthical and poare used for damage asseonomic dama

ife that is includ

ctural Measures

reflect the vacan sometimewhile deprecilife. Old goo

ductive and b

replacemenot apply; suritual significRB), and formher standards

e of money oe a higher oand their will

re usually basmages for thand amountease and damation densitie

nversion of f other damand general herature are bri

et part of staments6, this sh

in terms of eniform apprones apply difolitical questother damaessments preages.

ded in the stand

and Flood Proof

lue of a goodes overestimaation accouods which arbetter perfor

t cost accoch as artisticance tied to ts of cultural os of valuation

over time. Forr lower priceingness to co

sed on data ce current sitts of damagmage factors ces), while rel

damages intage types andealth impactsefly summari

ndardised mehould not ref

economic damoach given thfferent valuetions that areage categorieefer to assess

dard model for

fing

May 2010

d at the time ate damage, nts for the re damaged rming goods

ounting and c treasures, the belief of or biological n need to be

r example, a e next year onsume now

collected for tuation it is ges. Usually, can increase lative values

to economic d/or for the s are an easy ised.

ethods used flect on their

mage (e.g., a hat different es to human e not easily es. To avoid s the loss of

r flood

MRC Flood Ma

Guidelines Flo

3.6.2 Lo

In recent yJonkman, 2area: 1) Inform2) An ana3) An est The applicaflooding stuto estimatebreaches (shigh level oalso be useAn examplewhere the w

Figure 3‐5

For smallerassess humused to indperson’s instability in for maintaiwith a 1m dthe floodwa 3.6.3 H

Important nand their im

anagement and

ood Risk Assessm

oss of life

years several 2007). In gen

mation regardalysis of the eimate of the

ation of thesudied. For anye flood characsuch as in theof detail. Gend. In this anae of a mortaliwater is rising

Example of a

2007).

r‐scale floodsman instabilityicate the comnstability. Ovflows of relaning vertical depth and a vaters. Furthe

Health effects

non‐lethal hemportance is

Mitigation Prog

ment

methods haneral the follo

ding flood chaexposed popumortality (pe

se steps and y applicationcteristics (seee Netherlandneral estimatalysis the effety function isg rapidly, so t

a mortality func

s where velocy (i.e., ‘topplimbination of verall, the avtively low destability ran

velocity of 1 mr details on h

and social an

ealth impacts highly depen

gramme Compon

ave been devowing steps

aracteristics;ulation and eercentage of

the level ofn, the output e Section 3.3ds or New Ortes of the poects of warnis presented ithat people h

ction derived fo

cities are moing’) in flooddepth (h – [mvailable studepth‐velocity nge from 0.6 m/s would mhuman instab

nd environm

of floods arendent on (am

nent 2: Structur

‐ 81 ‐

veloped for tare needed

vacuation prdeaths) amo

f detail of thof hydrodyna). For large‐srleans), stepsssibilities forng systems an Figure 3‐5. have limited t

or large‐scale flo

ore important flows. In gem]) and velocdies and expmultiples. Thm2/s to aboost likely leadbility are inclu

ental impacts

e injuries andmong other fa

al Measures and

he estimatioto estimate

ocedures; ng the expos

e analysis deamic flood simcale flooding 2 and 3 canevacuation lso need to bThis functiontime to reach

ooding in the N

t, criteria havneral, depth‐city (v – [m/s]periments shhe critical deut 2 m2/s. Thd most individded in Appen

s

d illnesses. Thactors), gener

d Flood Proofing

on for loss ofloss of life i

sed populatio

epend on thmulations cag associated wn be applied wof the populbe taken inton is applicableh a safe place

Netherlands (Jo

ve been deve‐velocity mul]) that wouldhow that peepth‐velocity his means thduals to be sndix 4.

he occurrenceral health and

g

May 2010

f life (see n a given

on.

e type of n be used with dyke without a ation can o account. e to areas .

nkman,

eloped to tiples are lead to a ople lose multiples hat a flow wept into

e of these d medical

MRC Flood

Guidelines

infrastrufloodwatapproachcompreh In additiThere is (see, e.g. Flooding damageshiring of

3.7

The resuflood dam

The d

The categ

“’Unof th

Dam As the efinformat2.7). Thescope of Damage values) aper hectdistrict. Example A damagThis flooan extreinhabitanestimate Table 3‐1The spatshowing discussio

d Management a

Flood Risk Asse

cture in the ter (fresh wah to “predictihensive overv

on, there aralso evidenc., Bennet, 197

can also has are generalappropriate

Presentatio

ults of flood mage is calcu

damage amo

absolute damgories;

nderlying’ dathe overall dam

mage per hect

ffects of a flotion on the dae resolution othe analysis

per hectareare available tare over a la

e of a damage

ge assessmend scenario deeme coastal nts where floed with the D

1 offers an otial damage dthe density

ons on the rev

and Mitigation P

essment

area (beforeater or salt ing” these heview of the av

re often longce of a link b70).

ave adverse lly analysed bspecialists.

on of damag

damage assulated for a gi

ount for the fl

mage value b

ta (raw data mage) such a

tare as an ind

ood are spatamages. Geoof these mapand the data

can be detin a high levearger area w

e assessment

nt for a floodepicts the floflood. This sood consequutch standard

overview of tdistribution fof damage pvision of the

Programme Com

e and immedwater) and ealth effects. vailable infor

ger term psybetween psyc

effects on by means of

ge assessme

essment caniven flood sce

lood scenario

by category a

that has alres numbers of

icator of “da

ially distributgraphical infops and the le available.

ermined preel of detail pewhen damage

t

d scenario in oding of the scenario descuences can bdised damage

the distributifor a flood scper hectare. national polic

mponent 2: Struc

‐ 82 ‐

diately after duration of Hajat et al. (mation regar

ychological hechological he

the environf qualitative/d

nts results

n be presentenario, the fo

o (in USD or a

and the distri

ady been pref flooded hou

mage density

ted it can beormation sysevel of spatia

ecisely if the er unit area. e functions a

the Netherladensely popucribes the flbe serious. Te model at ro

on of the tocenario in theThese resultcy on flood p

ctural Measures

the flood), flooding, Th

(2004) and Arding health i

ealth effects ealth effects a

ment and thdescriptive m

ed in differeollowing pres

another curre

ibution of th

esented in muses, or flood

y” in the area

e relevant to tems (GIS) faal detail dep

underlying dIt is also posare available

ands is offereulated South‐ooding of anhe total damoughly 23.7 b

tal damage oe Netherlandts are used iprotection in t

and Flood Proof

climatic facthere is no shern et al. (2mpacts of flo

as a result and post floo

he social sysmethods and

ent ways. Assentations are

ency);

e total dama

onetary damed hectares o

a.

present mapcilitate this. (end on the (

data (e.g., lasible to averafor a larger

ed below as a‐Holland provn area with mage for thisbillion Euros.

over damageds is given byn flood risk the Netherlan

fing

May 2010

ors, type of standardised 2005) offer a oods.

of flooding. od mortality

stem. These require the

suming that e possible:

age over the

mages as part of farmland;

ps that offer (See Section (geographic)

and use and age damage area, e.g. a

an example. vince due to one million s scenario is

e categories. y Figure 3‐6, studies and nds.

MRC Flood Ma

Guidelines Flo

Table 3‐2

Residences

Commercial

Vehicles

Infrastructu

Airports

Agriculture

Direct dama

Total

Figure 3‐6

3.8 D

If no damagevents. Thethe type omeasurableclassificatiobenefits of The table bhowever, davailability approach h

anagement and

ood Risk Assessm

Distribut

Province in t

Damag

and inventory

l and public bu

re

ages due to bu

Damage inte

Data collectio

ge functions ae table belowof data and e items that on of flood dflooding shou

below can bedata on manyof secondaas been appl

Mitigation Prog

ment

tion of damage

the Netherlands

e Category

y

uildings

usiness interru

ensity per hecta

on

are available,w offers a gesources thacan be use

damages (seeuld also be in

e seen as thy of these itry data, timied in Stage 1

gramme Compon

e over damage c

s (Jonkman et

uption

are of the flood

, they need teneral overvieat could be ed as data. Te Section 3.ncluded.

he model listtems are stillme and budg1 of the FMM

nent 2: Structur

‐ 83 ‐

categories for a

al., 2008).

Direct

(mln

2

ing of South Ho

o be developew of the dautilised to The table is 2.1). As part

of informatl often impoget constrainMP‐C2.

al Measures and

a flood scenario

damage

n. EUR)

16,000

2,300

40

300

500

1,900

2,700

23,740

olland Province

ped using dataamage categocollect data,based on tht of this dat

ion for sevessible to colnts. Section

d Flood Proofing

o for South Holl

% of total d

6

1

1

10

e in the Netherl

a from historories, their d, and a sumhe presentedta collection,

ral items. Inlect given th3.9 shows

g

May 2010

land

damage

9%

0%

0%

1%

2%

8%

1%

0%

ands.

rical flood definition, mmary of d general , data on

practice, he limited how this



Table 3‐3 Overview of Flood Damage Data

Category

Definition

Type

Source of data

Summary of data item

s that could be collected

Direct da

mag

es

Loss of life and

injuries

Rep

orted nu

mbe

r of deaths, injuries and

missing

persons

due to the

floo

d

Second

ary

data

District/Provincial

level direct floo

d

damage repo

rts

‐ No. of d

eaths

‐ No. of injuries

‐ No. of m

issing

persons

Dam

age to

prop

erties and

infrastructure.

Estimated

rep

lacemen

t or restoration

cost of an affected

structure an

d inventories in the

targeted areas such as

houses, schoo

ls, offices, com

mercial and

indu

strial

buildings and

installation

s, hydraulic works, ene

rgy,

transportation

, infrastructures and

sup

plies/assets of

individu

als and bu

sine

sses

‐ No. of h

ouses collapsed or dam

aged

‐ No. of b

usiness collapsed

and

dam

aged

‐ Km

of roads

‐ No. of schoo

ls/classroom

s ‐ No. of irrigation sche

mes

‐ Other item

s. based

on the actual indicators tha

t are available in the

Provincial and

District damage

repo

rts

Crop

dam

ages

Agriculture land

inun

dated or agriculture produ

ction lost,

food

and

seed stocks dam

aged

‐ Ha of crops destroyed

‐ No. of fruit trees destroyed

‐ Ha of land

dam

aged

/covered

in sand

‐ Tonn

es of expe

cted

harvest lost

‐ Stocks of p

rodu

ce and

seeds lost

Livestock

damages

Drowne

d livestock and

loss of income du

e to distress

sales

‐ No. of d

ead po

ultry, large and sm

all livestock

Dam

ages to

fishe

ry

Dam

aged

aqu

aculture and

fishing eq

uipm

ent

‐ Fish/shrim

p po

nds damaged

‐ Fish cages, rafts and

traps dam

aged

/lost

‐ Boats lost

Dam

age to

padd

y

Inun

dation

dep

th‐dam

age relation

ships and effect of

flood

ing du

ration

and

tim

ing on

net ben

efits

Prim

ary da

ta

Focus Group

Discussion

‐ Hypothe

tical flood

dam

age to paddy

in relation to

inun

dation

dep

th and

tim

ing of floo

ds

Dam

age to

houses

Inun

dation

dep

th‐dam

age relation

ship and

hou

se value

Prim

ary da

ta

Hou

seho

ld/

busine

ss survey

‐ Actual and

hypothe

tical dam

age to hou

ses as

percen

tage of h

ouse value

in relation to

inun

dation

dep

th



Category

Definition

Type

Source of data

Summary of data items that could be collected

Indirect damages

Cost of illness

Costs associated with increased incidence of water borne

diseases compared to conditions of no flooding. Such

costs include additional cost of health care that families

and individuals spend on flood‐related illnesses; and

additional expenditures from local government for

disease prevention and treatm

ent and for sanitation

control during the flood

Primary data

District

authorities, line

agencies

‐ Amount spent by individuals on flood related

health care

‐ Loss of labour from flood‐related disease

‐ Amount spent on drugs, campaigns, care for

disease prevention

Cost of

temporary

relocation and

rescue

Costs by local government and NGOs incurred before,

during and after a flood

Primary data

District, NGOs

‐ Transport, temporary shelter costs, costs of food

and drinking water, etc.

Income losses

Losses due to disruption of economic activities and/or

services by the flood for (i) individuals, landless labourers,

families and enterprises; (ii) large commercial and

industry enterprises that have been partly or fully closed

during high flood water levels

Household survey

Business survey

‐ Cost of repair/reconstruction of residential

structures, including:

‐ Hypothetical damages to the residential

structures based on flood levels

‐ Agricultural assets/incomes, crop losses

‐ Livestock losses, fish/shrimp losses

‐ Other HH income sources losses (e.g., home‐

based businesses, hired labour)

‐ Loss of work days

‐ Health costs

‐ Damage to business structures, equipment and

inventory

‐ Losses in revenues and employment

‐ Private expenditures for flood protection

Higher cost of

living

Due to relocation water, food and fuel that otherw

ise

would be available on the farm

must be purchased

‐ Costs/purchases of food, water, fuel, etc.

Cost of

prevention

measure

Preventive m

easures

‐ Cost of sand bags, etc.

Cost of cleaning

up after the

flood

Cleaning of houses, shops and other buildings

‐ Labour costs and m

aterials

Benefits

Flood benefits

Positive aspects of flooding for agriculture, fisheries and

otherw

ise

Primary data

Focus Group

Discussion

‐ Household benefits from fishing the flood waters,

fertility and pest control savings. to fields.



3.9 Flood damage assessment in the Mekong River Basin

3.9.1 Scope and selected areas

In Stage 1 of the FMMP, several focal areas were selected to investigate in detail the flood damages and risks. These areas and their counties were: Country Focal Areas Thailand: Lower Nam Mae Kok Lao PDR: Bokeo province and the Lower Xe Bang Fai River basin Cambodia: Kratie province, Right Bank Bassac River and Left Bank Mekong River Viet Nam: Upper Se San River basin, Plain of Reeds (PoR) and Long Xuyen Quadrangle

(LXQ) The four delta focal areas in Cambodia and Viet Nam compose two transboundary areas, one east of the Mekong River and the other west of the Bassac River. The detail of the assessment is at district level, although part of the data is available at commune level. (See Section 3.9.4 on data collection and surveys, below.) Below, the general implementation of damage assessment is described and results are illustrated for three districts in Cambodia that were analysed as focal areas in the Stage 1. These are the districts of Koh Andet, Koh Thom and Kampong Trabek. More information on the damage assessments for these districts can be found in the Volume 2C “Flood Damages, Benefits and Flood Risk in Focal Areas”. 3.9.2 Approach and methodology

As described in Section 3.3.4, there are basically two approaches to flood damage/risk assessment: absolute damage and relative damage. In the absolute damage approach, historical damage data for an (administrative) area are used to assess the flood damage/risk in that area. In the relative damage approach inundation‐damage relationships are developed on a per unit (land area such as hectares or homes) basis, and the flood damage/risk is assessed by applying the per unit risk to the number of units in the concerned area. For the top‐down approach only a limited set of data is needed to determine flood damage (in absolute damage curves) and it should be available at government agencies (see Section 3.3.4). For the bottom‐up approach, various data sets and many processing steps are required to generate relative damage curves (see Section 3.3.4). In this study both approaches are applied for damages to housing and agriculture. For infrastructure damage and relief costs, only the top‐down approach can be applied due to a lack of adequate spatial information. Hydrological hazard data (peak water discharges and volumes for different return periods) are of sufficient accuracy to be used in risk assessment. The methodology can usually be applied basin‐wide. However, use of the data is not easy. Hydrological hazards must be converted into flood hazards (with measures of flooding depth and duration for different return periods). Limited topographical information and the lack of available high quality models hamper accurate flood damage/risk assessments. Flood damage/risk maps for LMB focal areas should, therefore, be considered with caution.



Flood vulnerability assessments have been made for a number of sample districts in the focal areas. These assessments allow for the estimate of indirect damages due to flooding in addition to reported direct damages. Damage curves could be estimated providing the relation between flood damages and inundation depth and time of flooding in certain areas. There is a GIS‐based tool that allows for the calculation of flood damages/risks by combining flood maps (see Chapter 2) and other available GIS data sets with their derived damage curves. This tool is, in principle, applicable basin‐wide, provided that it is supplemented with flood maps and that adequate GIS data sets are available. 3.9.3 Damage categories

In the LMB the most important damage categories that were analysed as part of Stage 1 of the project are:

Infrastructure and Relief The damages combined in this category included those to facilities, equipment and materials for education, health, irrigation, bank erosion, fisheries, transport, communication, industry, construction, drinking water and sanitation, along with the costs of rescue operations, support and relief.

Housing Housing damages measured the value of homes swept away or demolished by floodwaters, partly damaged or submerged, and those with damaged roofs, along with other private property damage such as damage to cultural and historical structures, offices, small industrial units, markets and commercial centres and warehouses.

Agriculture Agricultural losses were measured in rice growing areas, flower and vegetable growing areas, and areas for other annual crops and perennial crops, and also included losses to large and small livestock and poultry, agro‐chemicals and erosion of farm land and housing land.

The importance of these categories varies among the focal areas given their different topographical characteristics and pre‐existing levels of flood protection. The rank share of each category in terms of total flood damages for the areas in the four MRB countries are presented in Table 3‐4.

Table 3‐4 Observed rank share of damages in the focal areas

Categories Laos Thailand Cambodia Viet Nam

Public infrastructure 3 2 1 1

Housing 2 3 3 2

Agriculture 1 1 2 3

Source: Direct flood damage inventory

Description of rankings: Number 1 is the most important share; Number 2 is moderate share, and Number 3 is a small share.

3.9.4 Data collection and surveys

In Stage 1 of the FMMP a data collection and processing methodology was developed, which was expected to yield the required data for the focal areas, within the constraints of availability of secondary data, time and budgetary limitations. This data collection and processing methodology is summarised here.



Flood damage data is collected and flood level data is measured from various administrations at provincial, district and commune levels. At the provincial and district levels, flood damage records were collected for as many years they existed. Households and businesses were surveyed in order to inventory of flood vulnerability characteristics and direct and indirect flood damages. Focus group discussions in the project areas also provided insight into a range of issues such as the benefits of flooding, traditional coping mechanisms and community resilience. One of the purposes of surveys was to collect and/or validate existing information on direct damages for an agreed reference flood; and, to the extent possible, and to document and quantify indirect damages associated with the reference flood. For this purpose the 2006 flood was selected as this flood was considered the ‘reference flood’ given its characteristic as ‘average flood’. The year was appropriate since the survey was conducted while the memory of the flood was relatively fresh. Household and business interview respondents were asked to also estimate damages they would suffer if water levels were to reach higher levels than in 2006. The sources of data on direct and indirect damages included interviews with officials, households, and businesses:

Provincial and District officials Review and validation of existing data on casualties and direct damages caused by historic flood/erosion events. The data consisted of extensive listings of in‐kind and monetary damages. Damage categories included educational facilities, healthcare facilities and assets, road and transport facilities, irrigation facilities, water supply and sanitation infrastructure, power facilities, and communication infrastructure.

Households Collection of data from individual households for the reference year in which significant flooding/erosion occurred (2006). Data collection included household characteristics, characteristics of the recent flood(s), losses to residential structures, hypothetical damages to the residential structures for higher floodwater levels, agricultural assets and incomes, measures of crop losses, livestock losses, fish/shrimp losses, other non‐agricultural household income sources (e.g., home‐based businesses, outside employment) and health impacts.

Local businesses (non‐home‐based) Collection of data from individual businesses for the particular reference year in which significant flooding/erosion occurred (2006). Data collection included business data, characteristics of the recent flood(s), direct and indirect damages to business structures, equipment and inventory, impact on revenues and employment, and expenditures for flood protection.

In addition, the field work in each study area included collection of baseline vulnerability data and focus group discussions on preparedness:

Collection of baseline flood vulnerability data In consultation with local district and/or commune officials and existing databases, data were assembled to document key aspects of flood and erosion vulnerability. The data included: demographic information on population composition and growth, number of



households, household assets, economic activities, poverty statistics, land ownership information, land use and cropping patterns, fisheries production, animal husbandry, and inventories of the types of structures and infrastructure in the area. In addition, the baseline incorporated data on business activities, educational facilities and assets, healthcare facilities and assets, road and transport facilities, irrigation facilities, water supply and sanitation infrastructure, power facilities, and communication infrastructure. This data feeds readily into the flood data database in addition to being useful for the formulation of flood risk management strategies.

Focus group discussions on community preparedness The survey team facilitated focus groups with local leaders, community/mass organisations and men and women living in the study area who perform different economic activities and are of different socio‐economic status. The discussions elicited qualitative information on the beneficial and detrimental effects of floods and erosion. Prior to commencing the discussions, a guidebook was prepared to standardise the approach and to promote comparisons. The focus group discussions covered the history of flooding in each area, participants’ perceptions of ‘good’ and ‘bad’ floods, benefits of floods for paddy cultivation and fishing, traditional coping mechanisms and strengths and weaknesses of community preparedness and resilience to flooding.

For each of these approaches a survey tool was developed. The tools consisted of questionnaires (for the first three approaches) and a guideline (for the focus group discussions). For focal areas where bank erosion is the main cause of damages, a revised methodology in order to collect data at district level for each relevant river bank section, without the need for household or business interviews or focus group discussions. The full set of tools consists of:

Questionnaires for a Provincial and District Level Inventory of direct flood damages/bank erosion;

Questionnaires for District Flood vulnerability/bank erosion baseline database;

Questionnaires for District flood events/bank erosion;

Questionnaires for District Indirect flood/bank erosion costs;

Guide for Focal Group discussion;

Household questionnaire;

Business questionnaire; and

National Report outline. These tools were translated into local languages and spreadsheets for questionnaire analysis were prepared and provided to the survey teams, along with guidelines on how to prepare a national report. The data from the surveys, interviews, group discussions, and field observations, combined with the data in the governmental damage reports, serve as the basis for the damage functions and damage‐probability curves.



Overview of data sources for damage assessment using a GIS For the absolute approach to flood damage (see Section 3.3.4) the data requirements are lower than for the relative approach. A Geographic Information System (GIS) is required for the relative approach, combining spatial data with tabular data. Figure 3‐7 shows the difference schematically for agriculture, housing and infrastructure damage and risk assessments. Table 3‐5 offers an overview of the data sources used to determine flood damages in the Cambodian focal areas. Maps were produced at both district and commune levels, although the commune level data for housing damages were partly based on district‐level information (average house values).

Table 3‐5 Data sources used for mapping damages in the Cambodian Mekong Delta.

Category Definition Source

Information on flood characteristics: Hydrology and Topography

ISIS maximum water levels for different return periods of flooding

Max. water levels for different seasonal

periods (e.g., early flood season, rainy season)

and various return periods (e.g., once every 2,

5, 10, 25 or 100 years)

ISIS hydraulic model (MRC)

Digital Elevation Model (DTM, DEM)

Topographic data for the whole MRB MRC GIS data

Information on land use and values

District and commune/village boundaries

‐ MRC GIS data

Village locations and number of households per village

‐ MRC GIS data

Paddy coverage (Japan International Cooperation Agency (JICA) land use map)

Coverage (areas) of early floods and of

seasonal rice and rainy season rice cultivation

MRC GIS data, JICA

Average house values per district ‐ FMMP‐C2 surveys

Damage functions

‘Damage ‐ floodwater depth’ curves for damages to housing and rice harvests at the district level

Relation (mathematical function) between

the local flood level (flooded areas) and the

damage on housing or rice cultures

FMMP‐C2 surveys



Figure 3‐7 The absolute and relative flood damage assessment approaches require different data sets.

Data

re

qu

ire

d

GIS

: lan

d us

e

= vi

llage

and

hou

seho

ld d

ata

Sur

vey:

hous

ing

data

/ val

ues

Sur

vey:

agric

ultu

re d

ata/

val

ues H

ydra

ulic

mod

el (I

SIS

) dat

a: c

alcu

late

d flo

od le

vels

Flo

odin

g re

ports

: dis

trict

/ com

mun

e da

mag

e da

ta

GIS

: lan

d us

e

= ag

ricul

ture

Ag

ricu

ltu

re

(ann

ual/

pere

nnia

l cro

ps,

lives

tock

, pou

ltry,

ferti

liser

s,

chem

ical

s, e

rosi

on e

tc. )

Ho

usin

g

(hou

ses,

offi

ces,

she

ds, c

ultu

ral

build

ings

, ind

ustri

al u

nits

,

mar

kets

, war

ehou

ses

etc.

)

Infr

astr

uctu

re &

re

lie

f

(for

edu

catio

n, h

ealth

, irr

igat

ion,

fishe

ries,

tran

spor

t, co

mm

uni-

catio

n, in

dust

ry, d

rinki

ng w

ater

/

sani

tatio

n, re

lief e

tc. )

GIS

: lan

d us

e

= in

frast

ruct

ure

data

Sur

vey:

infra

stru

ctur

e da

ta/ v

alue

s

Flo

odin

g re

ports

: cou

ntry

/ pro

vinc

e da

mag

e da

ta

GIS

bas

e da

ta: D

TM

(DE

M),

adm

inis

trativ

e bo

unda

ries

Hyd

rolo

gica

l dba

se: m

easu

red

flood

leve

ls (a

lso

mod

el d

ata

can

be u

sed)

Rela

tive d

am

age (

%)

- flo

od d

epth

gra

phs

Absolu

te d

am

age (

$)

- flo

od le

vel g

raphs

co

mp

lex

sim

ple



3.9.5 Derivation of damage functions

For the estimation of damages for specific floods, the following data sources and equations need to be combined (also see Section 3.3.2 for similar information): a) Data on land use and values; b) Data on flood characteristics; and c) Damage functions. During the first stage of the study (in 2008) household survey were carried out to investigate flooding extents and flood damages in several districts. This survey showed that for some items (e.g., houses, roads) the maximum annual flood depth largely determines the amount of damages. In agriculture there is a difference in damages between the ‘early flood’ crop and the ‘rainy season’ crop. Damage to the ‘early flood’ crop is mainly caused by June – July floods, while damage to the ‘rainy season’ paddies is the result of October flooding. Government data from the commune level and MRC survey flood damage data taken from three Cambodian districts can be used to derive the relationships between floodwater levels and flood damages for the three categories of infrastructure and relief, housing, and agriculture. Figure 3‐8 shows the results for Koh Andet District.



Figure 3‐8 District flood damage curves for three categories of damages in Koh Andet District in Cambodia.

Damage functions for housing from the household and business survey In addition to the damage curves derived from the districts data, damage functions were derived using data from the household and business survey that relates damage to floodwater depth in and around houses. This approach followed the “relative damage approach” (see

Koh Andeth: Direct & indirect damages Infrastructure & Relief

0.0

0.2

0.4

0.6

0.8

3.0 3.5 4.0 4.5 5.0

max water level at Borey Cholsar (M+MSL)

Dam

age

(Mill

ion

US

D)

damage

fit

Koh Andeth: Direct & indirect damages Housing

0.00

0.01

0.02

0.03

0.04

3.0 3.5 4.0 4.5 5.0


Dam

age

(Mill

ion

US

D)

damage

fit

Koh Andeth: Damages Agriculture

0.00

0.40

0.80

1.20

1.60

3.0 3.5 4.0 4.5 5.0


Dam

age

(Mill

ion

US

D)

damage

fit



Section 3.3.4). Thus, functions include water depth and damage fraction. Figure 3‐9 shows the relative damage function derived for housing damages for Koh Andet District.

Figure 3‐9 Example of a damage function for housing, relating housing damage (as % of the total value) to

floodwater depths.

Damage functions for rice paddy cultivation Based on information collected during the Focal Group Discussions, damage curves for rice crops in relation to flooding depths in the districts of Koh Thom, Koh Andet and Kampong Trabek have been estimated both for early flood season rice paddies and for rainy season paddies. The results are presented in the graphs below, showing three levels of damage, depending on the timing of the flood. In this case, damages were most directly dependent on the time of year in which the flooding occurred. For other regions, damage to rice paddy cultivation can also be dependent on the duration of the flooding.

House damage curve - Koh Andet District (Takeo)

y = 0.0351e0.9444x

R2 = 0.9881

0%

10%

20%

30%

40%

50%

60%

70%

0.50 1.00 1.50 2.00 2.50 3.00 3.50

Flood water depth in yard (m)

Dam

ages

Temporary

Semi-temporary

Overall



Figure 3‐10 Relative flood damage curves for rice paddies, Cambodia.

3.9.6 Creation of damage maps

There are essentially two types of damage maps: 1. Damage maps based on absolute damage curves; and 2. Damage maps based on relative damage curves. Examples of both and explanations of the differences are presented in Section 3.3.4. Figure 3‐11 depicts the steps for creating the (absolute or relative damage curve based) maps. The data for creating the damage maps can be combined in several ways, but no matter which analysis is used, the (absolute or relative damage curve based) maps should be similar. 1. Damage maps based on absolute damage curves Absolute damage curves are created with historic flood damage data, relating

floodwater depths for a specific area to damage (see Figure 3‐8). The flood level data at the time/date of flooding can be taken from measurements at a gauge station in a district or determined using a hydraulic model covering that same district. Absolute damage curves are linked to the long‐term flood levels from the same station, generating in a damage probability curve. Damage figures for certain flood return periods can be interpolated from this curve. In the example in Figure 3‐11, the map shows district boundaries plus the damages for specific flood return periods in the district with the data represented on a ‘pie’ chart on one map and a bar chart on the other.

2. Damage maps based on relative damage curves Relative damage curves are created with damage figures related to the flood depths at

the damage locations. (For example, see Figure 3‐9 (housing) and Figure 3‐10 (rice paddies).) The damage data may come from a field survey or may be based on professional estimates. A land use map depicts the spatial distribution of different land use areas (such as rice growing areas). A flood map (see Chapter 2) shows the water depths in an area for a certain flood return period (based on hydraulic modelling

Damage Curve for Early Flood Paddy, Cambodia

0

100

200

300

400

500

600

700

800

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

Flooding depth (m)

Dam

ag

e (

US

$/h

a)

1 of June

1 of July

1 of August



results). By combining the land use map (e.g., rice paddies, villages or roads) with the flood map (water depths at specific flood return periods), the areas with rice paddies and certain water depths for different flood return periods are specified. This information can be combined with the relative damage curves, yielding a measure of the damages from the flooding of areas like paddies. Maps like those in Figure 3‐11 shows the damages for specific flood return periods in the area with the data represented on a ‘pie’ chart on one map and a bar chart on the other. Maps can also be generated by shading such as those in Figure 3‐12 and Figure 3‐13.

3.9.7 Results of damage assessments

The damage assessments results for district level focal areas in Cambodia have been generated using the absolute damage approach. Table 3‐6 shows the results for Koh Andet District for four floods of four different probabilities. The results can also be shown on maps. Figure 3‐12 and Figure 3‐13, for example, display the total damage and the damage per hectare for rice paddies.

Table 3‐6 Damage data for different return periods for Koh Andet District. Damage is shown in million

USD.

Probability (1/year)

Damage Category 1% 5% 10% 50%

Housing 0.05 0.04 0.03 0.00

Infrastructure and relief 1.14 0.75 0.69 0.10

Agriculture 2.15 1.58 1.48 0.35

Total 3.35 2.37 2.20 0.45

By combining the above damage assessments with the probability estimates, the overall risk can be assessed, as will be shown in Section 4.5. As the ‘relative damage curve’ method uses spatial information available on maps, the level of detail that may be obtained is higher than with the ‘absolute damage curve’ method. In both cases care needs to be taken if data from different detail levels is combined. To create maps based on relative damage curves, there is a need to perform a series of calculations both in a spreadsheet and in a GIS. For details on the steps and calculations in the spreadsheet, see the spreadsheet with the necessary guidelines that is included on the CD/DVD available at the MRCS FMMP Office in Phnom Penh.

MRC Flood Ma

Guidelines Flo

Figure 3‐11:

anagement and

ood Risk Assessm

Steps (simpli

Mitigation Prog

ment

ified) to create

gramme Compon

maps based on

nent 2: Structur

‐ 97 ‐

n absolute and

al Measures and

relative damag

d Flood Proofing

ge – flood depth

g

May 2010

h curves.

MRC Flood

Guidelines

Textb

Camb The IlocatiBy tawith dpossibtransfmap. Camb To caInterncan bto be be ‘raperioand bsee FThesedistric

d Management a

Flood Risk Asse

box 2: Creatin

bodia, creatin

SIS‐based floion maps, to king floodwadistrict‐averable to calculferring the d

bodia, creatin

alculate damnational Coope used as a b‘early flood’ ainy season’ d) with theseby combiningigure 3‐9) in e results can cts or commu

and Mitigation P

essment

ng Specialised

ng a housing d

oodwater or generate wa

ater damage age home vallate the flooata from the

ng a rice padd

ages to rice peration Agebasis. The padpaddies, whpaddies. Bye rice paddy the data witthe spreadshbe transferreunes (see Figu

Programme Com

d Damage M

damage map

basin wateater depth incurves for dues and dataod damage te spreadsheet

dy damage m

harvests forency (JICA) laddies that areile the other y combining maps in a Gth the rice daheet, it is posed to a GIS toure 3‐13).

mponent 2: Struc

‐ 98 ‐

Maps Based on

based on rel

r depth mapnformation (bdamage to hoa for the numo housing pt to a GIS, th

ap based on

r different flond use map e flooded oncpaddies (on the floodwa

GIS, transferriamage curvesssible to calcuo display dam

ctural Measures

n Relative Da

lative damag

ps can be coby return peromes (Figurember of houseer village anhese results c

relative dam

ooding returshowing thece every twohigher grounter depth ming the GIS ds (of e.g. Julyulate the dammages to the r

and Flood Proof

amage Curve

e curves

ombined witriod) for eache 3‐9), combieholds per vilnd per commcan be displa

age curves

n periods, thrice cultivati years are conds) are consimaps (by floodata to a sprey for ‘early flomage to the rrice crop on a

fing

May 2010

es

th village h village. ining this lage, it is mune. By yed on a

he Japan ion areas onsidered idered to od return eadsheet ood’ rice; rice crop. a map by

MRC Flood Ma

Guidelines Flo

Figure 3‐12

anagement and

ood Risk Assessm

Example of a

fields in 3 dis

Mitigation Prog

ment

a damage map

stricts and their

gramme Compon

based on relati

r communes in

nent 2: Structur

‐ 99 ‐

ve damage cur

Cambodia.

al Measures and

ves, showing to

d Flood Proofing

otal damage to

g

May 2010

rice paddy

MRC Flood

Guidelines

Figure 3‐1

d Management a

Flood Risk Asse

3 Example o

fields per

and Mitigation P

essment

of a damage m

r hectare in 3 di

Programme Com

ap based on re

istricts in Camb

mponent 2: Struc

‐ 100 ‐

elative damage

bodia and their

ctural Measures

curves, showin

communes.

and Flood Proof

g damages to r

fing

May 2010

rice paddy

CHA

APTEER 44

FLOOOD RISSK ASSSESSMENT

MRC Flood Ma

Guidelines Flo

4 F

4.1 In

In risk assechapter depresenting assessmentThe final se

4.2 G

Flood risk is Ris For examplexpected dyear. Since referred to In flood risassessed. Fincludes the 1. The st

This ihydrol

2. Next, shouldrelatiohazard

3. To assof flooassess

4. By com Figure 4‐1 each of thepicture, eac In Section 4

anagement and

ood Risk Assessm

LOOD RISK

ntroduction

essment, thescribes the athe assessmt can be usedection demon

General appr

s generally de

sk = Probabilit

e, if an area amage from this risk numas yearly exp

k assessmenFigure 4‐1 she following st

arting point nformation logical procesthe relationsd be derivedonship. Thesed assessmentess the floodod damage isment). mbining these

shows clevere steps yields ch graph shar

4.4 this appro

Mitigation Prog

ment

ASSESSME

results of bapproach toent results (Sd to evaluatestrates how

roach

efined by the

ty x Damage

has a probabsuch floodin

mber reflects pected econo

t, both the phows conceptteps:

is informatiocan be obtsses. ship betweed. This can e first two s. d damage a rs needed; i.e

e steps the re

rly, using seva set of datares a common

oach is elabor

gramme Compon

ENT

both hazard risk assessmSection 4.3). e the effectsrisk is determ

e simple equa

bility of floodng is USD20 the average

omic damage

probabilities tually the pr

on on the fretained from

en the floodwbe done ussteps are des

relationship be., the ’dama

eturn period o

veral graphs a that fits inton data axis w

rated for area

nent 2: Structur

‐ 103 ‐

and damagement (SectionSection 4.4 d of measuremined for a ca

ation,

ding of 1% pemillion, the rexpected ec.

and consequrocess of risk

equency of cehistorical d

water dischasing a hydrascribed in m

between the age function’

of a certain d

(starting in to the next st

with a graph u

as in the LMB

al Measures and

e assessmentn 4.2) and thdescribes hows and the levase study are

er year (or 1/risk would eqonomic dama

uences of flok determinat

ertain river fldata or pro

rge and theaulic model more detail in

flood charac(see Chapte

damage level

the lower rigep in an intesed in the pr

B.

d Flood Proofing

ts are combihe several opw the resultsvel of accepta for the LMB

(4.

/100 per yearqual USD 200age per year

ood events hation for a riv

floodwater diobabilistic an

e water (flooor a stage‐n Chapter 2

cteristics ander 3 on flood

can be asses

ght hand coregrated procerevious step.

g

May 2010

ned. This ptions for s of a risk table risk. B.

.1)

r) and the 0,000 per , it is also

ave to be ver and it

ischarges. nalysis of

od) depth discharge on flood

the level d damage

ssed.

rner) how ess. In the

MRC Flood

Guidelines

Damag

4) D

Figure 4‐1

4.3

4.3.1

To estimconsequenumber with a Mthat a repoints usfor all ev Howevera limitedfloods. Fevents aeach of tdamage

Table 4‐1

Return

Probabi

(per yea

Flood da

Differentrecurrendamage.

d Management a

Flood Risk Asse

3) Damagassessme

Dama

ge D

Determinatio

Conceptu

Presentatio

Economic r

mate the floences of flooof flood even

Monte Carlo segression anasing a randomvents and can

r, rather thand number ofFor example, re used as tthese flood rassessment.

Illust

period (year)

lity of exceed

ar)

amage (USD)

t flooding scece interval b Another me

and Mitigation P

essment

ge ent

age D

Retperio

on of the risk

ual approach fo

on of the res

isk

ood risk of od events thants should besimulation. Tlysis suggestm selection. In be combine

n looking at ef potential flothe annual fthe standard eturn period

rative summary

ance/flooding

enarios can bbetween two easure often

Programme Com

Water depth

urn od T

r determination

sults of risk a

an area, inft could affecte analysed wihe simulatios for existingn this way, fld to estimate

every possibleoods which flood, the 1/2return perios, models ca

y of damage an

2

g 1/2

Damage

e defined byfloods with used is the

mponent 2: Struc

‐ 104 ‐

Return perioT

Water depth h h

n of flood risk.

assessments

formation ist that area. Fith numericaln takes existg data to randood hazards e the risk.

e type of flooare represen2, 1/5, 1/10 od or frequen be used to

nd probability c

5

1/

e1/2 Dama

their return a certain flooe probability

ctural Measures

od

s

s needed onFor a full probl models. Oneting data anddomly generaand damage

od, the standntative for thand 1/100 (rency measureo generate a

alculations for

5

/5

age1/5 Dam

period or freodwater leveof ‘exceedan

and Flood Proof

Dischar

1) Probaanalysisdischarg

Dischar

2) Hydramodelin

n the probababilistic anae useful way d then uses tate data for mvalues can b

ard approachhe spectrum read: once pees I flood esthydraulic sim

a range of flood

10

1/10

mage1/10 D

quency; i.e., el and the conce’ of an e

fing

May 2010

rge Q

abilistic s of river ges

rge Q

raulic ng

abilities and lysis, a large to do this is he equation missing data be estimated

h is to select of possible er 100) year timates. For mulation and

d scenarios.

100

1/100

Damage1/100

the average orresponding event with a



specified damage; i.e., the probability (per year) that a certain damage value will be exceeded. These two metrics can be stated in terms of each other as follows: Probability of exceedance = 1 / return period (4.2) This means that the event with a 100 year return period has a probability of exceedance (flooding of greater magnitude and damage) of 1/100 per year. The relationship between the return period and the damage can be plotted in a graph (see Figure 4‐2 – left). The longer the return period, the more extreme the event and the higher the damages. However, even with higher return periods, damages will not increase exponentially. In fact, the expected damages begin to reach a plateau (the curve will ‘flatten’) as the maximum potential damage is reached and additional water adds little more to significant harm. The relationship between the probability of flooding and a given damage value can also be displayed graphically (see Figure 4‐2 ‐ right). Events that occur with a low probability and large damages are shown on the left side of this graph. The figures below are based on the assumption that damages only occur for floods that occur (on average) once per year or with a smaller probability. However, it is also possible that damages occur multiple times per year in the same area. In that case, a certain set of damages could be exceeded with a frequency that is larger than once per year. For example a certain damage value can be exceeded twice per year if such floods occur twice or more per year. In that case the right side graph in Figure 4‐2 could display frequencies/return periods that are larger than one. The data on damages is collected for each flooding event and the events data note the number of such floods.

Damage ($) Damage ($)

Return period (year)

Probability of exceedance

(1/year)

1 2 5 10 25 100 1/100 1/10 1/5 1/2 1

Figure 4‐2 Relationship between return period and damage (left) and probability of flooding and damage

(right).

Using this information, it is possible to determine the expected value of damages for an area, as presented below. The unit of risk is measured in annual monetary damages and is referred to as expected economic damage. Note that the risk/expected damage can also be determined for a single unit of land7; e.g., per hectare. In such case the risk is presented in terms of expected monetary damages per year per land area such as in USD/ha/year. The risk for a whole area is relevant when one wants to know the absolute level of risk and consider risk reduction measures (see also Section 4.4). The risk per hectare is of interest when one wants to compare relative risk between areas.

7 It is preferable to estimate the damage per hectare for the actual surface of land that has been flooded. However, the exact size of the flooded area is not always known and the damage needs to be averaged over a larger administrative area than only the flooded area.

MRC Flood

Guidelines

The risk/shows thfigure thalso the arisk (if thintegral c Note thalead to dmultiplyiway to tadditionaapplies to

Da

mag

e [U

S$/

ha]

1/100

B

b

A

Figure 4‐3

When a likely appjust sumomitted. The genperiods) (e.g. 100 The resudamage they offcontributlargest ceasier toprotectioincluded 4.3.2

An impoestimate

d Management a

Flood Risk Asse

/expected dahe relationshe risk can beareas ‘a’, ‘b’, he curve cancalculus).

at just multipdouble counting Damage1/the y‐axis, oal risk from eo the other d

1/25 1/10

C

c

The area

approxim

large numbeproach a cont

mming the ar

eral experieform a majo

0 years) can b

ults of risk cafor differentfer insight ition of specicontribution o see the ron works as in Section 4.

Risk to life

rtant indicatoe this type of

and Mitigation P

essment

amage can bhip between measured b‘c’, ‘d’ and ‘e

nnot be descr

plying the damting and an o

/2 with a retuoverlapping peach flood (pdamages/pro

1/5

D

d

below the curv

mated by summ

er of events wtinuous curveeas of blocks

nce is that tor part of thee very high, t

alculation ca areas can bn the spatiaific areas to to the overaelationships well as poin.5.

or in risk evarisk, mortali

Programme Com

be determineprobability oy summing ae’ can be appribed by an e

mages from Toverestimatiourn period of partly A, B, Cprobability ofbabilities.

e

ve, the total risk

ing the areas of

with differente. In that cases A, B, C, D

the events we risk. Althouthis effect is o

an be showne shown. Theal/geographicthe overall fall risk level between flo

nt to other i

aluation and ty data and r

mponent 2: Struc

‐ 106 ‐

ed by calculaof flooding areas A, B, C,

proximated byequation for

Table 4‐1 witon of the damf ½ per year oC and D. Insf 1/2 – proba

E

k or expected d

f the rectangles

t return perioe the risk/exand E and t

with the higugh damagesoften largely

n on maps. Fe advantage c distributioflood risks. Tcan be easilooding and nsights. Exam

decision‐makrelated flood

ctural Measures

ating the areand the damD and E. For y triangles an use of a mo

th their probmage. Figureone would ‘estead, one sability of 1/5)

1/2

damage (for exa

s, A, B, b, C, c, D

ods is knownpected valuethe triangles

ghest probabs for events wcompensate

For example of the use o

on of flood This implies ty identified. the presencmples of risk

king is the risd event data

and Flood Proof

ea under themage (Figure a more prec

nd added to fiore exact me

ability of flooe 4‐3 shows wextend’ the ahould calcula) x Damage1/

Probability

ample, in USD/y

D, d, E and e.

, the stepwiscan be approa, b, c, d an

ilities (or lowwith high retd by the low

the expecteof flood risk mrisks and that ‘hot spoMaps can a

ce or absenck maps for t

sk to life (losare required.

fing

May 2010

e curve that 4‐3). In the

cise estimate ind the total easure using

oding would why: by just rea E all the ate just the

/2. The same

y [-]

year), can be

se curve will oximated by nd e can be

west return turn periods probability.

ed economic maps is that thus to the ots’ with the also make it ce of flood he LMB are

ss of life). To . The loss of

MRC Flood Ma

Guidelines Flo

life can bedesigned to(similar to T The risk to shows the pat a logarith4‐4 shows a

Probexce

10-3

10-5

10-7

Figure 4‐4

In additionindividual rrisks of traf4.3 for furth

4.4 R

4.4.1 E

The results education m 1. Measur2. Measur Figure 4‐5 omeasures othe damagewith larger

8 A FN curve ivalue. The difcurve.

anagement and

ood Risk Assessm

e estimated o show the pTable 4‐1).

life is generprobability ofhmic scale, man example o

bability oeedance

10

Example of a

, the risk of isk due to floffic or diseasher details).

Risk reductio

ffects of anti

of a flood rimeasures. Ba

res to reduceres to reduce

offers an oveon risk. It asse up to a protprobabilities

is very similar tfference is that

Mitigation Prog

ment

with the aprobability of

ally referred f flooding leameaning that of an FN curve

of e (1/yr)

0 10

an FN curve.

flooding to ooding can bee. These resu

n measures

‐flooding and

sk assessmensically there

e the probabie the consequ

rview of possumes that mtection level s (smaller ret

o the curve thathe damage va

gramme Compon

proaches deflooding and

to as societading to a cethe difference8.

00 10

the individue compared tults can be u

and risk eva

d flood prepa

nt could be vare two cate

lity of floodinuences of floo

sible measuremeasures to rewith a certaiturn periods)

at shows the pralue (in this cas

nent 2: Structur

‐ 107 ‐

scribed in Sd the loss of

al risk. It canrtain numberce between t

000

al (death of o individual rused as input

aluation

aration measu

very useful togories of floo

ng; oding.

es and schemeduce the prn probability are controll

obability of excse loss of life) is

al Measures and

ection 3.6. Alife for sever

n be shown ir of fatalities.two axis‐units

Fatalitie

a person) carisks for othes for decision

ures on risk

o evaluate thod risk reduct

matically showobability willof flooding. ed and do no

ceedance resultshown on the

d Flood Proofing

Again a tablral events or

in an FN curv. Both axes as is a factor 1

es

an be calculaer activities sun‐making (se

he effects of tion measure

ws the effectsl completely In other worot cause any

ting in a certainhorizontal axis

g

May 2010

e can be scenarios

ve, which are shown 10. Figure

ated. The uch as the ee Section

flood risk es:

s of these eliminate ds, floods y damage.

n damage of the FN

MRC Flood

Guidelines

For examarea up tMeasurebuilding period. M Note thathe consreallocat

Measur

‐ Streng

‐ River w

‐ Retent

Figure 4‐5

4.4.2

Cost‐benreductionother woon cost ‐Rowsell athe UK (M Investmewherevedistributimpacts, analysis. effects. In such acomparethe activbenefits managemstrengtheeconomidifferent

Damag

1

d Management a

Flood Risk Asse

mple, this is tto a certain lees to reduce timprovemenMeasures can

at while comsequences, thtion of the flo

res to reduce

gthening of dy

widening (“roo

tion and emer

Effects of

Basics of co

nefit analysisn. A project ords, the ben‐ benefit anaand ChattertMAFF, 1999,

ent projects er possible, gional effectsincluding inToo often, h

a ‘limited’ coed with the mvity is considare lower t

ment this mening of floc flood damat types of co

ge ($)

1/100 1/10

and Mitigation P

essment

he case whenevel; e.g., for the consequents, reduce thn be combine

pensation anhese measureood damages

flooding prob

ykes and emba

om for rivers”

rgency storag

f measures on f

ost‐benefit an

s is used to should be uefits generatlysis in the cton (1977) or2000) and Au

can be sysgiven a mone and uncertantangible onehowever, in p

ost ‐ benefit ameasurable cdered attracthan the comeans that od proofing)ages to detests have to b

P 1/5 1/2

Programme Com

n dykes or ema major flooences of floohe consequend, so there w

nd insurancees do not redcosts.

bability

ankments

”)

e

flood risk.

nalysis

evaluate whndertaken if ted by the prcontext of flor in the guideustralia (BTRE

stemically anetary value. ainties whilees, must bepractice, the

analysis, the costs of the active (it geneosts, the actcosts of i

) are comparmine whethbe included:

Probability of exceedance

(1/year)

mponent 2: Struc

‐ 108 ‐

mbankmentsd of 50 year oding such asnces of an evewill be a comb

are also ofteduce the tota

Measure

‐ Land us

‐ Second

‐ Flood p

above

hether a cerit results in oject should ood managemelines for cosE, 2001).

nalysed withCost ‐ bene

e pointing totaken into aanalysis is o

measurable activity. If therates an inivity is not ncreasing saared with thher an investcosts of inve

Damage

1/

ctural Measures

s are built to return periods adaptation ent (total dambined reducti

en mentionedal damages. T

es to reduce f

se planning

dary flood defe

proof building

ground levels

rtain measuran increase exceed its coment is givent‐benefit ana

h individual fit analysis p alternativesaccount in a often narrowe

economic bee benefits arcrease in ecattractive. Iafety againshe decrease ment is attraestment (fixe

e ($)

/100 1/10

and Flood Proof

provide protd. of land use cmages) for a gon in risks.

d as measureThey are only

flooding conse

ences

s (e.g. raising

s)

re achieves eof economicosts. Further in the workalysis (CBA) d

effects estimprovides an os. In theory, competent ed to tangibl

enefits of anre higher thaconomic benn the contest flooding in the yearactive. In the ed and variab

Proexc

(

1/5 1/2

fing

May 2010

tection to an

controls and given return

es to reduce y a financial

equences

buildings

enough risk c welfare. In information

k by Penning developed in

mated and, overview of all relevant cost‐benefit le monetary

n activity are an the costs, nefit). If the ext of flood (e.g., dyke

rly expected cost figure, ble) and the

obability of ceedance (1/year)



costs of maintenance and management. The benefit measure is the reduction of the expected economic damage. Restating this in terms of an equation, a flood risk reduction investment is attractive in the terms of a cost ‐ benefit analysis if:

)(DEI (4.3) where: I = investment in project/measure [e.g in USD/yr]

ΔE(D) = reduction of expected economic damages [e.g. in USD/yr] The Greek letter delta here is used to mean ‘the change’ but of course the change in expected economic damages here is best described as the reduction in the damages since the investments are designed to result in changes that reduce damages. The reduction of the risk/expected economic damages can be calculated with a simple equation looking at the before and after changes:

NDEDEDE )()()( 0 (4.4)

where: E(D)0 = expected economic damage in the original situation [e.g. in USD/yr]

E(D)N = expected economic damage after measures/project completion [e.g. in USD /yr]

The expected economic damage can be determined through the approaches outlined in Section 4.2. The change in the expected damage value due to different types of risk reduction measures can be determined according to the principles presented in Section 4.3.1. The above approach works if both the investments and the benefits can be expressed in monetary units for the same unit of time (e.g., USD/yr). In many cases, investments in flood risk reduction are one‐time costs where the benefits are expected to continue over a long period, such as for a dyke construction project, whereas the benefits are expressed on an annual basis. To conduct a cost benefit, the time element needs to be standardised for both the investment costs and the benefits. The way to standardise the benefits is to calculate their net present value. Any costs and benefits that occur over future time periods need to be discounted to determine their net present value. The following general formula (as presented in Appendix 3) can be used to do that:

n

t

ii

r

CBNPV

1 )'1( (4.5) where: NPV = net present value [e.g. USD]

Bi = benefits in year i [e.g. USD] Ci = costs in year I [e.g. USD] t = time [year] r’ = reduced interest rate [‐]

The reduced interest rate r’ is generally used for discounting. It is equal to the rate of economic growth minus the rate of inflation. When one assumes an infinite time horizon the discount factor approximates 1/r’ and the net present value of the benefits becomes ΔE(D)/r’.

MRC Flood

Guidelines

4.4.3

A specifideterminare depethat are dyke. The Deltasystems.following In econoreductionstrengtheto floodwprobabiliassociatebe plotteminimal.correspocorrespo This applevel forresults hNetherlaprobabilisafety sta

Figure 4‐6

Hig

d Management a

Flood Risk Asse

Economic o

c form of thene an optimaendent on thprotected by

a Committee The Commig a major floo

omic optimisan of the risk. en and raise waters exceeity of floodined damage. Bed as a functi This is consonding dyke honding protec

roach was ar the largest have been ands. For coaities betweenandards were

Principle

gh probability

and Mitigation P

essment

optimisation

e cost ‐ beneal level of prohe level of pry dykes. The

e in the Netheittee began iod that cause

ation, the incIn the examthe dykes. Teding the floong (the higheBy summing tion of the safsidered the height is knowction level in

pplied after flood proneused to destal areas, thn 1/4,000 pee set at 1/1,2

of economic op

Programme Com

fit analysis isotection wheotection. Thichoice of pr

erlands has aits work to ied enormous

cremental inple of the Nehis simple apod defences. er the dykesthe costs andfety level. A poptimal poinwn. Because tterms of a pr

the Dutch 19e area, Southrive safety he optimal ler year and 1/250 per year a

ptimisation.

mponent 2: Struc

‐ 110 ‐

s economic on both the inis is the caserotection leve

applied econonvestigate pdamage in th

vestments inetherlands, thpproach assuThus, each d

s the smallerd the expectepoint can be nt. (See Figuthe statistics robability can

953 storm suh Holland (vstandards fevel of prote/10,000 per yand 1/2,000 p

Small p

ctural Measures

optimisation. nvestments ine, for examplel is depende

omic optimisapossibilities fohe Netherlan

n more safethe investmenmes that floodyke height cr the probabed damage odetermined wre 4‐6.) At tof the water

n be defined.

urge to detevan Dantzig, or flood dection has beyear. For Dutper year.

probability

and Flood Proof

This approacn protection e, for flood ent on the h

ation to floodor new safetds in 1953.

y are balancnts consist of oding can onlcorresponds ility of floodr risk, the totwhere the totthis optimumr levels are al

rmine an op1956). Eventefences throen chosen wtch river basi

fing

May 2010

ch is used to and the risk prone areas height of the

d protection ty standards

ced with the the costs to ly occur due to a certain

ding) and an tal costs can otal costs are m point, the lso known, a

ptimal safety tually these oughout the with flooding in areas, the

MRC Flood Ma

Guidelines Flo

Economic oexample, thtaking into case of floo 4.4.4 R

The estimatto define ceSection 4.3Vrijling et a For individuactivity, suc10‐4 per yea For societal4‐7). The limthat any nuwith larger measures, tthe limit lisituation is

Acceptabl

Figure 4‐7

The above Netherlandfor risk to approach to 4.4.5 R

A tool used(vertical axand expectgreen), unasubjective damage cataken into a

Probabilexceeda

10-3

10-5

10-7

Calrisk

anagement and

ood Risk Assessm

optimisation he optimal leaccount the

oding and the

Risk valuation

tes of risk toertain standa3.2 for definial., 1998; Jonk

ual risk, a thch as a probaar) the situati

l risk, a limit cmit line sets tumber of fataconsequencthe calculatene, the situanot acceptab

e societal risk

Societal risk

concepts fos where newlife. At preso safety stand

Risk matrix

to evaluate is) and expeted damagesacceptable (redeterminatiotegories (ecoaccount.

ity of ance (1/yr)

10 100

Limit line

lculated k

Mitigation Prog

ment

can also be vel of raisingadditional co

e probabilities

for loss of lif

life can be iards that indiitions). Morekman et al., 2

reshold levebility of deation is not acc

can be deterthe acceptablities beyondes is acceptad risk can beation is acceble and addit

k

of fatalities due

or limits for w safety stansent, the Nedard for flood

risk is the riscted damages, a certain red) or subjectons regardingonomic dama

Fat1000

e

gramme Compon

applied to og a house to osts of raisins of several w

fe

ncluded in riicate the acce information2009), but a s

l can be set th of 10‐5 per ceptable and

mined througle number ofd this are conable only wite compared weptable. If thional measur

e to flooding an

individual anndards are deetherlands apding.

k matrix (seees (horizontarisk can be rt to further dg the risks. Iage, loss of

talities

nent 2: Structur

‐ 111 ‐

other types ominimise floong the buildinwater levels th

sk valuation eptable leven on this tosummary of t

to indicate tyear. If the arisk reductio

gh use of a lif fatalities vesidered unacth a smaller pwith the limit he calculatedres are neede

Unacceptabl

nd the limit line

nd societal reveloped thappears to be

e below). It prl axis). Basedranked as acdiscussion on In order to clife, environ

Probability exceedanc

1

10-3

10-5

10-7

Calculrisk

al Measures and

of problems od damages cng, the damahat would lea

and decisionl of individuapic is found he main conc

the tolerableactual calculan measures a

mit line in thersus the probcceptable. In oprobability. Iline. If the cad risk is aboed.

le societal ris

e of acceptable

risk are curreat meet sociee the only co

rovides a rand on the magcceptable (shits acceptabicategorise thmental or so

of e (1/yr)

0 100

Limit line

lated

d Flood Proofing

of various sccan be deterage to the strad to flooding

n‐making. It isal or societal in several t

cepts is given

e risk due to ted risk is higare necessary

e FN curve (sbability of a hother words,n planning palculated riskove the limit

sk

harm.

ently exploreety’s acceptaountry that

king of the pgnitude of phown on theility (orange) he damages, ocial impacts

Fatali1000

g

May 2010

cales. For rmined by ructure in g.

s possible risk. (See texts (see n below.

a certain gher (e.g., y.

see Figure harm such , an event protection k is below t line the

ed in the ble limits uses this

robability robability e chart in based on different s) can be

ities

MRC Flood

Guidelines

The choiassessmeis a subje

Probab

ility H

M

L

Figure 4‐8

4.5

4.5.1

This sectthe resulthe areasrelevant Figure 4‐time serprobabiligraph indamagesrelations(shown i

Figure 4‐9

4

4.5

5

5.5

6

6.5

7

7.5

8

8.5

1910

1914

1918

Flo

od

/ w

ate

r le

ve

l (m

+ M

SL

)

0

10

20

30

40

50

60

70

0

Dam

age

(% o

f house

val

ue)

d Management a

Flood Risk Asse

ce of the boent can only ective determ

High

Medium

Low

Structure

Risk assess

Introductio

tion presentslts of the floos under studyin an overall

‐9 outlines thies output ofity of differe the upper s (shown graship with then the graph i

Approach

Hydrolo

1922

1926

1930

1934

1938

1942

1946

1950

Dam

0.5 1 1.5

Flood / water

and Mitigation P

essment

undaries betprovide informination.

Low

e of a risk matrix

sment for th

n and approa

s the results od hazard any. Note that brisk valuatio

he approach f the ISIS hydnt water levright). The raphically in e flood probn the lower r

h for estimation

ogical data

1954

1958

1962

1966

1970

1974

1978

1982

Year

mage

2 2.5

depth (m + field level)

Programme Com

ween the catrmation on th

Medi

Consequ

x.

e LMB

ach

of risk assessalyses (Chapbenefits of flon.

that was fodraulic modeels can be aelationship bthe lower lbability data,right of the fig

n of the risk.

1986

1990

1994

1998

2002

2006

3 3.5

mponent 2: Struc

‐ 112 ‐

tegories is anhe level of ris

ium H

uences

sment withinpter 2) and daooding are no

llowed in thel (the graphnalysed and between flooeft) is also , the damaggure).

1

1.5

2

2.5

3

3.5

1/10

0

1/5

0

Flo

od

/ w

ater

dep

th(m

+ f

ield

leve

l)

0

200

400

600

800

1000

1200

1400

1600

1800

2000

1/10

0

1/5

0

Dam

age

($ /

ha)

ctural Measures

n arbitrary onsk and not on

High

n the contextamage assessot discussed i

is project to in the uppepresented godwater leveknown. By

ge probability

Flood dept

1/1

0

1/5

P

1/1

0

1/5

P

and Flood Proof

ne. The outcon the accepta

t of the MRBsments (Chapin detail here

estimate riser left of the raphically (shels (or flood combining ty curve can

h - Probability

1/2

Probability (-)

Risk

1/2

Probability (-)

fing

May 2010

omes of risk bility, which

B; combining pter 3) from e, but can be

sk. From the figure), the hown in the depth) and the damage be derived

MRC Flood Ma

Guidelines Flo

The resultsareas in theand Kampo 4.5.2 R

As part of tsimulated. flood evengenerated. return perioThe Koh Tu

Figure 4‐10

The risk (aestimating Table 4‐2 shFrom these

Table 4‐2

District

Koh Andet

Koh Thom

Kampong T

The results for the avereach circle

anagement and

ood Risk Assessm

s obtained use FMMP Stagong Trabek.

Results of risk

the risk assesThe damage nts. By comFrom the figods; especialm district has

Probability d

nnual expecthe area unhows both the results it bec

Results o

Trabek

of the risk carage flood risshows the m

Mitigation Prog

ment

sing this appe 1 are illustr

assessment

ssment the flocurves frombining thesegure it can bly for eventss the lowest o

damage curve fo

ted value ofder the probhe absolute excomes clear t

of risk calculatio

Expected e

(mill

alculations cask calculationmagnitude o

gramme Compon

proach for thrated below.

oods that ocm Section 3.8 e results, thbe seen that s with a relatoverall risk cu

or three Cambo

f the damagbability damxpected valuthat Koh And

ons for the thre

economic dam

ion USD/yr)

0.77

0.42

0.69

an also be visns for the focof the risk fo

nent 2: Structur

‐ 113 ‐

hree districtsThese are th

curred betwewere used the followingKoh Andet htively large purve.

odian districts.

ges) for the age curve. (Se and the expdet has the hi

ee focal areas i

mages Exp

sualised on mal areas in Caor the district

al Measures and

in Cambodie districts of

een the yearso estimate tg probability has the higherobability of

three districtSee explanatpected value ghest risk.

n Cambodia.

pected econo

(million U

0

0

0

maps. Figure 4ambodia andt. The distrib

d Flood Proofing

a designatedKoh Andet, K

s 1910 and 2he damages y damage cuest damages flooding (40

ts can be detion in Sectio per square k

omic damages

USD/(km2*yr)

0.0022

0.0009

0.0014

4‐11 shows thd Viet Nam. Tbution over t

g

May 2010

d as focal Koh Thom

006 were for these urve was for most % ‐ 60%).

erived by on 4.3.1). kilometre.

s per ha

)

he results he size of the three

MRC Flood

Guidelines

damage is clear thand Kamshare of

Figure 4‐1

The figurthe fact t

d Management a

Flood Risk Asse

categories (hhat the distri

mpong Trabekthe risk is to

1 Flood risk

re shows cleathat the Vietn

and Mitigation P

essment

housing, infrabution over tk, the largestinfrastructur

ks by district for

arly that the namese area

Programme Com

astructure andthe damage ct part of the re.

r selected areas

risks on the s are more de

mponent 2: Struc

‐ 114 ‐

d relief, agriccategories va risk is to ag

s in Cambodia a

Vietnamese ensely popul

ctural Measures

culture) is shoaries betweengriculture. Fo

and Viet Nam.

side are muated.

and Flood Proof

own within en districts. For Koh Thom,

ch higher. Th

fing

May 2010

each circle. It or Koh Andet , the largest

his is due to

CHA

APTEER 55

REFFERENNCES

MRC Flood Ma

Guidelines Flo

5 R

[1] Abt S.Rzone,Water [2] AdamsoBasin in ChiManageme [3] Ahern Mfloods: epid [4] Anonymsummary. [5] Bennet,Community [6] BTRE (2Transport aPrinters. Ca [7] ClausenPlanning. UPolytechnic [8] Dutta, DHydrology, [9] FEMA (2Federal Em [10] Hajat consequencEnvironmen [11] Hoskinmoments. C [12] JonkmaUniversity. [13] JonkmEconomic M [14] JonkmAmerican W [15] Jonkmstudy for dy [16] Karvonbreak analy

anagement and

ood Risk Assessm

REFERENCES

R., Wittler R.Jr Resources B

on, P.T. (2007iang Rai Provent Programm

M., Kovats Sdemiologic ev

mous (2009), D

, G.(1970) By Disaster, Bri

2001). Econoand Regionaanberra: Com

, L.K. (1989) Unpublished c collaboratin

D., Herath S., 277:24 –49.

2003). HAZUSergency Man

S., Ebi K.L., ces of floodintal Science a

g, J.R.M. andCambridge U

an S.N. (2007

an S.N., BocModelling of F

an S.N., PenWater Resour

an S.N., Vrijliyke ring Sout

nen R.A., Hepysis, RESCDAM

Mitigation Prog

ment

S

J., Taylor A.,Bulletin Vol. 2

). Strengthenince, Thailan

me, MRC, Sep

.R., Wilkinsovidence, Epid

Development

ristol Floodsitish Medical

omic Costs oal Economicsmonwealth o

Potential DamM.Phil. thesig with Binnie

Musiake K. (2

S: Multi‐hazarnagement Age

Kovats S., Mng in Europeand Public He

d J.R. Wallis (1niversity Pres

7) Loss of life

karjova M., KFlood Damag

ning‐Rowsellces Associatio

ing J.K., Kok h Holland, Ris

ojoki A., HuhM Final Repor

gramme Compon

, Love D.J. (25 No.4, pp. 8

ning of the Flod. Hydrologicptember.

n P., Few R.demiologic Re

t and use of D

s 1968 – ConJournal Vol.

of Natural Ds, Disaster Mof Australia.

m Failure: Esis at the Sche and Partner

2003). A mat

rd loss estimaency, Washin

Menne B., Edwe and the imealth Vol 1 No

1997). Regionss, Cambridge

estimation in

Kok M., Bernge in the Neth

l E. (2008) Hon (JAWRA) V

M. (2008) Floisk Analysis V

hta H.K., Louhrt, Helsinki U

nent 2: Structur

‐ 117 ‐

1989) Huma881‐890.

ood Managemcal Baseline S

, Matthies Feviews Vol. 27

DACA in de lo

ntrolled Surv3, pp.454‐45

Disasters in Mitigation Re

stimation of Chool of Geogrs, Redhill.

hematical flo

ation model mngton.

wards S., Haplications foo. 1 pp. 13‐21

nal frequencye, U.K.

n flood risk as

nardini P. (20herlands, Eco

uman instabVol. 44 No. 4

ood risk asseVol. 28, No. 5,

hio A. (2000) niversity of T

al Measures and

n stability in

ment capacitStudy. Inform

. (2005) Glo7 pp. 36‐46.

ower Mekong

vey of Effect8.

Australia. Reesearch Wor

Consequencesgraphy and P

ood loss estim

methodology

ines A. (200r public heal1.

y analysis: an

ssessment, P

008) Integratelogical Econo

bility in floodAugust 1008

ssment in th pp.1357‐137

The use of phTechnology, H

d Flood Proofing

n a high floo

ty for the Kokmation and Kn

bal health im

g basin – exec

ts on Health

eport 103, Brking Group,

s, and ImplicPlanning at M

mation, Journ

y – flood mod

03) The humath, Journal o

approach ba

hD thesis, De

ed Hydrodynomics 66, pp.

d flows, Journ8, pp. 1‐11.

e Netherland73.

hysical modelHelsinki, Finla

g

May 2010

od hazard

k River nowledge

mpacts of

cutive

of Local

Bureau of Paragon

ations for Middlesex

al of

del.

an health of Applied

ased on L‐

elft

namic and 77‐90.

nal of the

ds: A case

ls in dam‐and.



[17] Karvonen R.A., Hepojoki A., Huhta H.K., Louhio A. (2000) The use of physical models in dam‐break analysis, RESCDAM Final Report, Helsinki University of Technology, Helsinki, Finland. [18] Kelman I. (2002) Physical vulnerability of residential properties in coastal, eastern England, PhD dissertation, Cambridge University. [19] Kelman, I. and R. Spence (2004) An Overview of Flood Actions on Buildings. Engineering Geology 73. pp. 297–309. [20] Kok M., Huizinga H.J., Vrouwenvelder A.C.W.M., van den Braak W.E.W. (2005) Standaardmethode2005 schade en slachtoffers als gevolg van overstromingen, HKV report PR999.10 ‐ draft translation is available as: Standard method for Predicting Damage and Casualties as a Result of Floods. [21] Leenders, J.K., Roelevink, A., Parodi, G. and J. Wagemaker (2009a). Development and Demonstration of DACA (Damage and Casualties Assessment) in the Chiang Rai Region, Lelystad: HKV Consultants and ITC. [22] MAFF (1999). Flood and Coastal Project Appraisal Guidance. Economic Appraisal (eds. K. Riddell and C. Green). London, UK: Ministry of Agriculture, Fisheries and Food, Flood and Coastal Defense Emergencies Division. [23] MAFF (2000). National Appraisal of Assets from Flooding and Coastal Erosion. Halcrow Maritime, HR Wallingford, Flood Hazard Research Centre for the Ministry of Agriculture, Fisheries and Food, Flood and Coastal Defence Emergencies Division, London, UK. [24] Merz B., Kreibich H., Thieken A., Schmidtke R. (2004) Estimation uncertainty of direct monetary flood damage to buildings, Natural Hazards and earth System Sciences Vol. 4, pp. 153 – 163 [25] Messner, F., Meyer, V. (2005): Flood damage, vulnerability and risk perceptions; HKV: Implementation of DACA. [26] Messner, F., Penning‐Rowsell, E., Green, C., Meyer, V., Tunstall, S., Veen, A. van der (2007). Evaluating Flood Damages: Guidance and Recommendations on Principles and Methods. FLOODsite Consortium, Wallingford, UK. [27] Parker D.J., Green C.H., Thompson P.M.,(1987) Urban flood protection benefits – a project appraisal guide. Gower Technical press. [28] Penning‐Rowsell, E.C., Chatterton, J.B., (1977). The benefits of flood alleviation – a manual of assessment techniques, Saxon House, ISBN 0566001908. [29] Roos, W. (2003) Damage to buildings. Delft Cluster report DC1‐233‐9. [30] Royal Haskoning (2008) Flood damages and flood risks in focal areas. Annex 2 to the draft Stage 1 evaluation report, (August 2008). [31] Stuyt L.C.P.M., Reinders J.E.A., van der Hoek E.E., Hermans A.G.M., Muinck Keizer de M., Kamperveen J.P., Hoogvliet M.C., Gerritsen A.L., Hounjet M.W.A., Bontenbal M.P., Hoefsloot F.V.J., Icke J. (2003) Environmental impact of flooding, Delft Cluster Report DC1‐233‐8.



[32] Thieken, A., Muller, M., Kreibich, H., Merz, B. (2005). Flood damage and influencing factors: new insights from the August 2002 flood in Germany, Water Resources Research, 41:1‐16. USBR (United States Bureau of Reclamation) (2001) Downstream hazard classification guidelines. Acer Technical Memorandum no. 11. [33] Van Dantzig D. (1956) Economic decision problems for flood prevention, Econometrica Vol. 24 pp. 276‐287. [34] Van der Veen A., Steenge A.E., Bockarjova M., Logtmeijer C.J.J. (2003) Structural economic effects of large scale inundation: a simulation of the Krimpen breakage, Delft Cluster report DC1‐233‐12. [35] Viessman, W., G.L. Lewis and J.W. Knapp (1989). Introduction to Hydrology. Third edition, Harper Collins Publishers. [36] Vrijling J.K., van Gelder P.H.A.J.M. (2002) Probabilistic design in hydraulic engineering. TU Delft lecture note WA5310. [37] Vrijling, J.K., Hengel, W. van, Houben, R.J. (1998) Acceptable risk as a basis for design, Reliability Engineering and System Safety Vol. 59 pp. 141‐150. [38] Watkins, L.H. and D. Fiddes (1984). Highway and urban hydrology in the tropics. Pentech Press Ltd, London.

APPPEENDDICEES

MRC Flood Ma

Guidelines Floo

Appendix 1 This Appenguidelines f(EXCIMAP) follows theintroductio In 2007, the2007/60/ECmember stprepare floplans (by 20 The Flood Dto other reg

Membwithinestimaenviropossibhydrolretentinfrastlong‐teoccurr

In the States,betwe

Flood flood cnaturaplanni

Flood on presystemor sub

Flood practicareas i

In the shall nrisks uunlessamong

Membthe flo

Membreview

9 http://eur‐le

nagement and M

od Risk Assessme

1 Approa

dix presents for its implemon behalf of e EU Directivn to the man

e European PC9 on the assetates of the ood hazard a015).

Directive intrgions or river

ber States sha their territoates of the ponment, cultuble issues suclogical and ion areas, tructure, the erm developrence of flood

case of an in, Member Sten the conce

risk managemconveyance ral floodplainng, land use,

risk managemevention, proms. They shou‐basin.

risk manageces, improvein the case of

interests of snot include mupstream or these measg the Membe

ber States shood hazard m

ber States shaw and updatin

ex.europa.eu/L

Mitigation Program

nt

ach to Flood

the Europeanmentation proseveral Europve, and the agement of f

Parliament anessment and European Und flood risk

roduces a numr basins. Thes

all, for each riory, undertakpotential adveural heritagech as the togeo‐morphothe effectiposition of

pments, andds.

nternational rtates shall eerned compet

ment plans sroutes and ars, environmnature conse

ment plans sotection, and uld also take

ement plans ment of watf a flood even

solidarity, floomeasures whidownstreamsures have ber States conc

all make avaaps, the floo

all encourageng of the floo

LexUriServ/LexU

mme Component

Damage an

n Union’s Flooduced by thpean countriDutch Damaflood damage

nd the Councmanagemen

Union to (preks maps (by

mber of key se principles i

iver basin or ke a preliminerse consequ and econompography, thological charveness of populated a

d estimates

river basin orensure that etent authorit

shall take intoreas which h

mental objecervation, nav

shall address preparedneinto account

may also inter retentionnt.

od risk manaich, by their

m of other cobeen coordincerned.

ailable to thed risk maps a

e active involvd risk manag

UriServ.do?uri=

t 2: Structural Me

App. 1 ‐ 1

d Risk Asses

od Directive he European es, the FLOOage and Case and flood ri

il of the Euront of flood riseliminarily) a2013), and t

principles thinclude the fo

portion of annary flood risuences of futmic activity.he position oracteristics existing hu

areas, areas of the imp

r sub‐basin wexchange of ties.

o account coave the potectives, soil avigation and p

all aspects oss; including t the characte

nclude the p, as well as

gement planextent and iountries in thated and an

e public the pand the flood

vement of ingement plans

=CELEX:32007L0

easures and Flood

ssment in Eu

along with pExchange CirDsite researcualties Assessk in Europe.

pean Union pks. This direcssess their fo prepare flo

at may be reollowing:

n internationask assessmenture floods fThese shouldof watercoursincluding floman develoof economic pacts of clim

which is sharerelevant inf

osts and beneential to retaiand water port infrastru

of flood risk mflood forecaeristics of the

romotion of the controlle

s establishedmpact, signifhe same rive agreed solu

preliminary frisk manage

terested part.

0060:EN:NOT

d Proofing

urope

reliminary strcle on Floodch project thassment (DAC.

promulgatedctive compellflood risks (bood risk man

elevant and a

al river sub‐bnt along withfor human hed consider ases and theioodplains asoped flood c activity andmate change

ed with otherformation ta

efits, flood exin flood watemanagementucture.

managementasts and earlye particular r

sustainable ed flooding o

d in one Memficantly increer basin or sution has be

flood risk assement plans.

ties in the pr

May 2010

udies and Mapping at directly CA) as an

Directive ed the 27 by 2011), nagement

applicable

basin lying h damage ealth, the as best as ir general s natural defence

d planned e on the

r Member kes place

xtent and er such as t, spatial

t focusing y warning iver basin

land use of certain

mber State ease flood sub‐basin, en found

sessment,

oduction,


Guidelines Flood Risk Assessment App. 1 ‐ 2 May 2010

EXCIMAP In order to carry out the tasks set by this directive, a large number of European countries organised themselves into the European Exchange Circle on Flood Mapping (EXCIMAP). In 2007 they produced the Handbook on Good Practice for Flood Mapping in Europe10, which focuses on the initial steps required by the Directive. Flood maps are used to indicate the geographical areas which could be covered by flood waters from all natural sources at several levels of probability. The Guideline distinguishes between: (i) river flooding in floodplains; (ii) sea water flooding; (iii) mountain torrent activity or rapid run‐off from hills; (vi) flash floods in Mediterranean ephemeral water courses (‘wadi’s’); (v) groundwater flooding; and (vi) lake flooding. EXCIMAP flood maps can show flood extent according to: (i) probability classifications; (ii) historic events; (iii) flood depths; (iv) flow velocities; (v) flood propagation; and (vi) degree of danger. These maps can be used for: (i) land use planning and management; (ii) water shed management; (iii) hazard assessment; (iv) emergency planning and management; (v) planning of technical measures; and for (vi) overall public awareness. For flood risk management, flood maps are useful tools in answering questions such as: (i) where is the greatest risk? (ii) where should investments be targeted? (iii) which investments may have the highest cost effectiveness? Flood hazard maps and related maps EXCIMAP directs that flood hazard maps be prepared for low, medium and high probabilities of flood occurrence. For each probability the extent, inundation depth/floodwater level, and where appropriate the water velocity are indicated. These maps are to be prepared using hydraulic models, statistical analysis and observations. Flood danger maps indicate a degree of danger by combining information on probability level, flood depth, velocity and debris content. These maps are not required by the Flood Directive, but are used in some member countries for land use planning. Flood event maps record historical flooding and can be seen as a first approach to flood hazard mapping. The can be used in awareness raining, for follow‐up of flood hazard assessment, for the calibration of models, and as an emergency management and planning tool. Flood vulnerability maps Flood vulnerability maps show the categories and assets at risk: (i) population, (ii) assets and economic activity, and (iii) potentially affected installations that may release pollution. In the Netherlands, these maps are prepared at the provincial level and are available to the public and specialised services11.

10 http://ec.europa.eu/environment/water/flood_risk/flood_atlas/index.htm 11 www.risicokaart.nl



Flood risk maps The EU Flood Directive defines ‘flood risk’ as the combination of the probability of a flood event and of the potential adverse consequences for human health, the environment, cultural heritage and economic activity associated with a flood event. The basis of flood risk maps, therefore, is to indicate: (i) the number of inhabitants potentially affected, (ii) the type of economic activities potentially affected, (iii) installations which might create incidental pollution, and (iv) other pertinent information, such as high content of transported sediments and debris and other significant sources of pollution. Risk may be calculated as: Risk = C * Ph where: C = potential adverse consequence and

Ph = the probability of the hazardous process Risk is expressed as a potential loss in a particular area or unit area (ha) within a given time period (one year) using a second equation: C = V * S * E where: V, S and E are called the ‘vulnerability parameters’ for damage measurement: V = value of element at risk (in monetary terms or human life) S = susceptibility: damage effect on element at risk (as a function of

magnitude of hazard (e.g., floodwater depth‐damage and damage‐duration relationship curves (with each factor ranging from 0 to 1)

E = exposure: the probability of the element at risk to be present while the event occurs (factor ranging from 0 to 1) FLOODsite12 was a research project on Flood Risk Analysis and Management Methodologies, designed by the EU to support its Member States in complying with the requirements of the EU Flood Directive, collecting data during 2004‐09. FLOODsite research analyses the physical, environmental, ecological and socio‐economic impacts of floods from rivers, estuaries and the sea. It considers flood risk as a combination of hazard sources, pathways and the consequences of flooding on the “receptors” – people, property and the environment. The FLOODsite research consortium included 37 of Europe’s leading institutes and universities and the project hired managers, researchers and practitioners from a range of government, commercial and research organisations, specialising in aspects of flood risk management. FLOODsite research is currently applied in EU Member States in the form of models and technical support for river, estuary and coastal sites in Belgium, the Czech Republic, France, Germany, Hungary, Italy, the Netherlands, Spain, and the UK. FLOODsite also prepared a discussion paper13 reviewing the flood damage evaluation methods applied in the UK, the Netherlands, the Czech Republic and selected German States, followed by a guideline for flood damage assessment14.

12 www.floodsite.net 13 Meyer, V.(2005): National flood damage evaluation methods

14 Messner, F. et al. (2007): Evaluating flood damages: guidance and recommendations on principles and methods



Dutch Damage and Casualties Assessment Tool15 In the Netherlands, (HIS‐SSM), a standard method for flood damage and casualties assessment has been used since its development in 1999. It is used to assess the change of risk following national and regional decisions on water management measures in the Netherlands. Since 2007, regulations require that all Dutch water‐projects of national importance be evaluated using this tool. It is also used in the production of the Dutch flood vulnerability maps mentioned in Section 3.7. The Damage and Casualties Assessment Tool (DACA) is a GIS‐based application in which land use, inundation depth‐damage functions per land use and maximum damage values of each land use are combined with a model‐generated flood hazard. This is in line with the ‘relative approach’ described in Section 3.3.4. These approaches are an advance on traditional flood measures and preparations. Traditionally, flood planning focused on safety standards, such as dyke design levels or reservoir volumes required to ensure predefined protection levels. Protection of the community against floods occurring with a frequency of once in 1,250 years is an example of such an established level under the flood protection law of the Netherlands. What is missing in this approach is the amount of damage protection that the system offers. In other words, the approach disregarded the efficiency of flood protection measures. While economic costs of alternative flood defence options are usually considered in the decision‐making process, the benefits of flood protection in the form of prevented damages should be taken into account too. The new paradigm for flood risk management includes the economic analysis of costs and benefits of flood protection and mitigation measures. Not only the safety of a defence system and its associated costs are considered, but also the damages to be expected in case of its failure. As a consequence of the application of cost‐benefit and risk analysis, safety standards can better be adjusted to the specific circumstances, given that the costs of ensuring an overall safety standard often considerably exceed the benefits in some areas.

15 Based on: Messner, F. and Meyer, V. (2005)

MRC Flood Ma

Guidelines Floo

Appendix 2 Flow velociloads. Sevecombinatiocriterion frosimulationsexpected dand Kelmancoastal areprobabilistican lead to failure of wimpacts of consideredpractical vatests.

Figure A 1

In general tflow velocitneeded to r The degreeFor examplMeasures aof structuranot be direconducting

nagement and M

od Risk Assessme

2 Approa

ity is an impoeral models hon of flood deom brick ands, such criteridamage. Othen and Spencas and addec model for partial or ful

walls. The modwaves and . A general walidation. Mo

Damage crite

et al., 2000)

the advantagty) are taken relate the loc

e of damage le wooden hand constructal damage. Tctly transferra detailed as

Mitigation Program

nt

aches for the

ortant factorhave been deepth and flowd masonry hoa can be useer researchee (2004) inveed the influecollapse of bl collapse of del of Roos copounding deweakness of ost of the pre

erion for brick a

).

e of the abovinto accountcal flood cond

will also be hhouses have tion techniquThis is why ered to other ssessment of

mme Component

e Assessmen

r in damage eveloped thaw velocity. Aouses (Clauseed to developrs have inveestigated phnce of rise rbuildings, wha building. Thonsiders hydebris. Also ththese ‘collapesented rela

and masonry h

ve approach t. A related diditions to the

highly dependifferent damues, e.g. buildxisting moderegions or tobuilding type

t 2: Structural Me

App. 2 ‐ 1

nt of Structu

to building st relate the

As an exampln, 1989). In cp flood risk mestigated buiysical vulnerrate. Roos (2hich takes inthese are 1) throstatic and e strength (rpse of buildintionships are

ouses as propo

is that the acisadvantage i vulnerability

ndent on the mage patternding houses oels and structo use with difes and their v

easures and Flood

ral Damage

since it impacphysical dame Figure A 1 combination maps that exhlding vulneraability of res2003) developto account twhe scour of fohydrodynamresistance) ongs’ models ce based on t

osed by Clausen

ctual flood cos that a lot oy of an individ

type of struns than concon poles, will tural damagefferent housivulnerability.

d Proofing

to Buildings

cts the hydromage to houspresents thewith input frhibit differentability. Kelmasidential proped a compwo failure mooundations, aic loads, as wof the buildinconcerns thetheory or sca

n (figure from K

onditions (flooof detailed (Gdual building

cture that is crete or bricalso affect the relationshiping types wit

May 2010

s

odynamic ses to the e damage rom flood t levels of an (2002) perties in rehensive odes that and 2) the well as the ng type is eir limited ale model

Karvonen

od depth, IS) data is .

affected. k houses. he degree ps should thout first

MRC Flood Ma

Guidelines Floo

Appendix 3 There are from a persfinancial buperspectiveintangible gapply if calc Several pricomprehenprinciples asections. Thexaminatioessential to Economic p

Economic adamages towho would The analystreflect a vieopportunityreduces nacapita weal In an econoof a good overestimainvolves imsubstituted Apply econ

The economrelatively econsidered market gooideal compdistortions.are necessaThis meansof transfer competitiveaccount for Time value

The time vathis year orvalue” in thwould rathe

nagement and M

od Risk Assessme

3 Fundam

different ratspective of a urden that ie and seek togoods and seculations of fl

inciples of ensive flood daas regards thhe application of flood dao follow these

points of view

aspects of flo the nationareceive the b

t has to poinew point of ty cost are uational per cath is a “bene

omic damage at the tim

tion of damamprovements by new, mor

nomic prices f

mic valuationasy to accomto identify th

od correspondpetitive mark Since these ary in order , real world mpayments like market pricr social benef

of money

alue of moner holding “mohe future refer save or ha

Mitigation Program

nt

mental issue

ionales for esingle personindividual oro assess the imervices. This lood damage

economic vaamage valuathe economic on of the prinamage, and te principles as

w

ood damageal economy. Tbenefits (if an

nt out the flothe society, tused in estimapita wealthfit”.

study, depreme when it age from a brs: old housere productive

for valuation

n of goods anmplish. Howevhe ”correct“ vds to its scarcket with maideal conditito calculate market priceske taxes and sces as well as fits and losses

ey is the wayoney” and wflects how pve something

mme Component

es in Econom

economic van or firm, negr firm has tompact on natbroader eco

e are to be us

aluation neetion study in valuation of

nciples of ecthe use of thes closely as p

e valuation aThere is no mny) as result f

ood damagesthe economicmating floodh is a “dama

eciated valueis damagedroader econos/goods whie and better p

n of damages

nd services wver, accordinvalue. As stacity price. Anany competinions are nevethe ideal “shs must be trasubsidies andby adjustings (externalitie

y of thinking waiting to purpeople generg they will us

t 2: Structural Me

App. 3 ‐ 1

mic Valuation

luation. Finaglecting publio bear. Econtional or regionomic persped for public

d to be coa consistent f flood damaonomic valueir results in practicable.

and analysis matter of whfrom the floo

s in a view pc and/or shad damages age” and any

s should be a by a floodomic perspecich are damperforming o

which are trang to economted in most end a scarcity ng actors iner fulfilled inhadow pricenslated into sd by convertinfor the sociaes).

about the crchase those ally think abe.

easures and Flood

n

ncial valuatioc affairs and nomic valuatonal welfarepective is thpolicy.

nsidered in way. Some oage are specation is crucflood risk an

require a deo would payd.

oint of the sdow prices ofnd benefits.ything that in

applied in ordd. Using repctive, becausemaged duringones.

ded in the mmic theory seveconomic texprice emergevolved and a real‐world“ of the gooshadow priceng monopolisal value where

hoice betweegoods in a f

bout the futu

d Proofing

ons considerfocusing on ttions have ae, including ime appropriat

order to cof the most icified in the ial for an apnalysis. There

eterminationy for the dam

society as a wf goods and . Thus, anytncreases nat

der to reflect placement coe replacemeng a flood ar

market is a taveral rules nextbooks, the ves in the contwithout gov

d market, adjod under exaes by excludinstic market pe market valu

en holding refuture year. Ture and whe

May 2010

r damage the actual a broader mpacts on te one to

onduct a mportant following

ppropriate efore, It is

of flood mages and

whole. To services ‐ hing that tional per

the value ost is an nt usually re usually

ask that is eed to be value of a text of an vernment justments mination. ng all kind prices into ues fail to

eal goods The “time ther they



There is a simple way to think about this. If you have USD100 today and lend it to someone or put it in a bank, you expect to be paid interest for the use of the money (for example, 10% per year). This means that from USD100 today you will have USD110 one year later. In another words, the value of USD100 today worth USD110 next year and can buy 10% more (or 10% minus inflation). In economic analysis the economist compares value of money at a different time instead of the amount of money. To do this, economists use a ‘discounting’ technique for converting value of money to a specific time/date. A number of terms, presented below, can describe this relationship over time, using mathematical equations. Compounding factor This factor is used to determine future value (FV) of a present amount (PV) at the end of the nth period at the interest rate of i: FV = PV (1 + i)n An easy way to think about the compound factor is to think of the ‘compound interest rate’ from bank interest, with the same interest multiple applied to the new amount at the end of a period of time, such as every year. In the example above for USD100 in the bank, if the compound rate is 10%, applied every year, at the end of 2 years, you would have, USD121. Compounding factor of once per year with annual distributions This is used to calculate the future accumulated value (FAV) at the end of nth period at an interest rate of i with a sequence of equal payments (A). (1 + i)n ‐ 1 FAV = A ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ i Here, the easy way to think about this is for a period (n) of 2 years, like above. In the example of USD100 in the bank, you would receive USD10 after the first year and USD10 after the second year. Of course, the total amount you would have would depend on what you do each year with the USD10 in interest. If you reinvest it at the same interest rate, then you have USD121 after 2 years as in the example above. But if you spend the interest each time you receive it, your future accumulated value is only USD20. Capital recovery factor This is used to calculate the amount of each payment (A) at the end of each n period to recover the present amount (PV) with the interest rate of i: i (1 + i)n A = PV ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ (1 + i)n ‐ 1 Net Present Value (NPV) Net Present Value is the present worth of the income stream (the continual annual payments, A, above, or the benefits or value of a project) generated by an investment. It can be calculated by the following equation:

NPV =

n

tt

tt

i

CB

1 )1(

)(



where: B = benefit; C = cost of investment and annual cost

T = time from year 1 to year n; and I = discount rate

Note that this equation doesn’t actually use profit but calculates an annual ‘net benefit’ of the project that is equal to the benefit minus the cost (B less C) in each year. The Greek letter sigma represents that this is a summation from year one to the final year, n. Constant price

The concept of constant price assures that a monetary sum taken from one year can be compared with a sum from another year. Flood damage evaluation is carried out in a specific year but the cost and damages are always estimated using data from previous years. To keep the values constant, the values from each year (called the ‘nominal value’ for that specific year) needs to be converted to a constant/fixed value using the current pricing value. To do this means applying the same idea of interest rates as applied in the equations above to the year of the measurement and bringing it up to the current year using information on comparative prices over the years. ‘Constant’ or ‘fixed price’ refer to a value from which the overall effect of general price changes has been removed. Using constant prices ensures that the past and future costs and benefits of the identified alternatives are estimated in the same units as the costs and benefits measured at the time of evaluation for decision.

MRC Flood Ma

Guidelines Floo

Appendix 4 For smallerpeople in fdrowning oused to asundertakenshows the close stabilitmultiples rabe related characteristinstability is

0

0,5

1

1,5

2

0

wat

er d

epth

(m

)

Figure A 2

Based on sfunction of above withdepth (in mdifferent leflood condi

nagement and M

od Risk Assessme

4 Analys

r‐scale floodsflood flows cor direct drowssess the flon (Abt et al.,combinationsty in flows wange from 0.to test circ

tics (weight, s given in Jon

0

Observation

velocity that

such tests crwater depth floodwater

meters or feetevels of dangtions.

Mitigation Program

nt

is of Human

where veloccan be an imwning (death)od risks at 1989; Karvos of depth anith limited de6 m2/s to abcumstances height, clot

nkman and Pe

1

flow

s from experim

resulted in ins

iteria have bh and flow vevelocity (in mt) on the y‐axger. The abov

mme Component

Instability i

cities are momportant rea). Criteria hava local leveonen et al., nd velocity thepth‐velocityout 2 m2/s. T(ground surfthing). Furthenning‐Rowse

2

w velocity (m/

mental series reg

tability.

been developelocity. An exameters or in xis. Distinctiove criteria ca

t 2: Structural Me

App. 4 ‐ 1

n Flood Flow

re importantason of leadve been devel. Different s2000). For thhat resulted y multiples. TThe differencface frictioner backgrouell, 2008.

3

s)

garding the com

ped that shoample of sucfeet per secoon is made ban be used to

easures and Flood

ws

t, loss of stabding to loss oeloped to assseries of exphe available in instability.he obtained ces in the rep, test confignd informat

4

mbination of wa

ow different h a criterion ond) on the etween threeo determine

d Proofing

bility (i.e. topof conscioussess this. Theperiments hatest data, F. It shows thcritical deptported test reguration) or ion regardin

data A

data R

hv=1,

hv=0,

hv=2m

ater depth and

levels of danis shown in tx‐axis and floe ‘zones’ reprisk zones fo

May 2010

ppling) for sness and se can be ave been igure A 1 at people h‐velocity esults can personal g human

Abt et al.

Rescdam

35m2/s

6m2/s

m2/s

flow

nger as a the figure oodwater presenting or certain

MRC Flood

Guidelines

Figure A 3

Finally, indiseases)directly directly das physinfrastru

Management an

Flood Risk Assess

Criteria fo

presents

ndirect flood) can also coestimating thdependent osical charactcture, demog

d Mitigation Prog

sment

or danger to pe

depth and velo

ing effects (entribute to lohe effects ofn the area thteristics (safgraphics, clim

gramme Compon

eople in flood fl

ocity both in me

e.g. on local foss of life. Hf these prochat is affectefety preparmate and ecol

nent 2: Structural

App. 4 ‐ 2

ows (USBR, 200

easures of mete

food supply) owever, theresses on theed and the soation, govelogy, etc.).

l Measures and F

01). The graph

ers and feet.

and longer tre are no que loss of lifeocial and econrnment and

lood Proofing

is two dimensio

erm health imantitative ap. Such lossesnomic conditd communit

May 2010

onal but

mpacts (e.g. pproaches to s should be tions as well ty self‐help

MRC Flood Ma

Guidelines Floo

Appendix 5 In order to projects int 1. Initiatio2. Plannin3. Produc4. Monito5. Closure A project stsome otherequipment planning/deproject plaimplementeproject impadjustmentare closed,

Figure A 4

The Best Developme a) Prelimb) Feasibc) Prelimd) Detailee) Constr The numbespecific prefeasibility st Each sectiowill be indimargins.

nagement and M

od Risk Assessme

5 The Be

facilitate mato phases suc

on ng/Developmtion/Implemoring/Controle

tarts with anr objective. Inand tools areevelopment/n is produceed; the prodplementationts. After comand the prod

The phases

Practice Guent/Design. Th

minary/pre‐feability study anminary design ed design andruction/bid d

er and conteneferences. Ttudy for engi

n of the guidicated by dis

Mitigation Program

nt

st Practice G

anagement oh as the follo

ent/Design entation

n idea to solvn the initiatioe purchased, design phaseed and the dduction take may reveal pletion, the duct or result

s of an enginee

uidelines arehis phase can

asibility studynd overall pla

d detailed plaocuments

nt of the stagehe preliminaneering proje

delines appliesplaying symb

mme Component

Guidelines an

of an engineeowing five ph

ve or mitigaton phase, fin and the ideae. The feasibdesign is pres place andthe need to project will bt may be hand

ring project.

e almost exn be subdivid

y nning

anning

es may differary design sects.

es to one or mbols that rep

t 2: Structural Me

App. 5 ‐ 1

nd Project P

ering project, ases:

e a problem,ances are moa begins to tability of the idpared. In Phthe projectcorrect the be closed; thded over to a

xclusively aped in various

r, depending stage, for ex

more of the apresent each

easures and Flood

hases/Stage

project man

, create a proobilised, a prake shape. Thdea is testedhase 3, the pis underwa

planning ande project teaa client.

pplicable to stages, listed

mainly on prxample, is o

above stages.stage (show

d Proofing

es

nagers norma

oduct or struroject team ihe second phd, and, if sucplans and deay. Monitorind/or design, aam disbands,

Phase 2: d below.

roject type oroften include

. In the guidewn below) in

May 2010

ally divide

ucture, or s formed, ase is the ccessful, a esigns are ng during and make accounts

Planning/

r country‐ed in the

elines this the page



Phase 2 contains all of the following five stages: a) Preliminary/pre‐feasibility study A pre‐feasibility study is the precursor to a feasibility and design study. Its main purpose is to decide whether it is worthwhile to proceed to the feasibility study stage and to ensure there is a sound basis for undertaking a feasibility study. A pre‐feasibility study generally includes:

Definition of achievable project outcomes;

Analysis of the development situation and constraints the project is to address based on collected data;

Identification of related (government and other stakeholder) policies, programs and activities;

Preliminary assessment of the viability of alternative approaches; and

Preliminary identification of likely risks to feasibility and benefits (including risks to sustainability).

b) Feasibility study and overall planning If a project is considered to be feasible based on the pre‐feasibility study, a more thorough feasibility study can begin. A feasibility study defines the project and its objectives in detail, and looks at these various forms of feasibility:

Technical feasibility: Can the measures technically be realised in the local context?

Operational feasibility: Are the measures to be implemented manageable by the local people?

Economic feasibility: Is the cost‐benefit analysis favourable?

Social feasibility: Are the objectives and measures socially acceptable?

Environmental feasibility: Are the environmental impacts acceptable?

Political feasibility: Will the measures be supported by those with power?

Overall feasibility: Will implementation of the envisaged measures result in accomplishment of the project objectives? Field surveys, hydrological and hydraulic analyses (in flood mitigation projects), social and environmental assessments, stakeholder meetings, and costs estimates are among the basis for answering the above questions. If the answers are positive, the operations/management structure and management method can be defined, and any initial planning can be detailed. c) Preliminary design

If a project is deemed feasible, the preliminary design stage can start. This stage focuses on the technical measures, as follows:

Site surveys and investigations and computer modelling provide the data for preliminary design criteria;

The design criteria can be translated into the preliminary design of structures and measures in an integrated and balanced system in which the envisaged management activities are linked to one another;

The cost‐benefit analysis (construction and operation) and analysis of environmental, social and political factors demonstrate whether the project is still viable.

If necessary, project planning can be adjusted based on new insights gained in this stage.



d) Detailed design and detailed planning During the final design stage, the detailed architectural and engineering drawings (the blueprints) of all physical components of the project are produced. Virtually all design problems must have been resolved before the end of the final design stage. Sufficient detail must be provided by the drawings and the report to allow reasonably accurate estimates of construction and operating costs, as well as the construction scheduling.

e) Construction documents/bid documents The detailed designs and construction scheduling are incorporated in construction documents and bid specifications, giving the contractors the information they need for construction. Where sections of the guidelines refer to other than the above‐described phases (e.g. the construction or monitoring phase), the following symbol is used.



Appendix 6 Flow Chart to Create Flood Maps with a GIS

METHODOLOGY TO CREATE FLOOD MAPS WITH ARCGIS BASED ON ISIS HYDROLOGICAL MODELING RESULTSLegend:

Notes: This diagram assumes the creation of separate layers for Cambodia and Viet Nam.

●

●

●●

To create water level or water depth maps on the basis of output of the hydrological model (ISIS), the full diagram needs to be followed.Is the ISIS node information already transferred to the GIS and are the water levels already included in the Access tabel (.mdb; see A-7), than only the lower grey section needs to be followed for mapping of a different level/ depth.

If the ISIS schematisation changes, the whole diagram has to be followed again.For water levels not yet included in Access, both grey areas need to be redone.

B-13

C-1

C-3

ISIS node names (ID's) and coordinates of

reservoir and cross-section nodes

(.csv)

ISIS node names (ID's) and water levels

(e.g. for return periods or scenarios)

(.csv)

ISIS polylines of river cross-sections

(.shp)

ISIS polygons of reservoirs

(.shp)

Import and cleanin Excel and transfer to

Access

Import and cleanin Excel and transfer to

Access

ISIS node names (ID's), type,

coordinates, location and w.l.'s in one table

(.mdb)

Combine in e.g. Access using queries

Add to ArcGIS project and clean

Add to ArcGIS project and clean

Import in ArcGIS, georeference and save in

shape-format (.shp)

Split in Reservoir and (River) Cross-section

points layers for Camb. & Vietn.

ISIS reservoir nodes (names, location and

w.l.'s) of Camb.(.shp)

ISIS cross-section nodes (names,

location and w.l.'s) of Camb.(.shp)

ISIS reservoir nodes (names, location and

w.l.'s) of Viet Nam(.shp)

Base data originating from

ISIS or MRC

Derived or created data by

FMMP-C2

Performed action on data

Draw 150 m buffer polygons around

polylines

Assign data of nodewithin polygon to polygon

('spatial join'); remove poly-gons with >1 node inside

Assign data of nodewithin buffer to buffer ('spatial join'); remove buffers with >1

node inside

Import and process in e.g. MatLab ('return

period' data)

Assign data of bufferto cross-section polylines;

remove polylines without node data

ISIS cross-section polylines in Viet Nam

with node ID and water levels(.shp)

ISIS cross-section polylines in Camb. with

node ID and water levels(.shp)

Convert polylines to points (e.g. 15 points/

polyline)

Convert polygons to points (e.g. 15 points /

polygon)

Put all points (original nodes + derived points)

in one layer

ISIS nodes, polyline points (and polygon points) in Viet Nam

with node ID and w.l.'s(.shp)

ISIS nodes, polyline points (and polygon

points) in Camb. with node ID and w.l.'s

(.shp)

Convert polygons to rasters using required

water level value

Create raster of requiredw.l. by interpolation (cell size as DTM); check ISIS whether

to use all points

Rectangular area (raster) containing

focal area in Camb. with w.l.'s

(.adf)

Rectangular area (raster) containing

focal area in Viet Nam with w.l.'s

(.adf)

Superimpose reservoir raster on point-based

raster (using 'con')

Remove the raster outside the focal areas

('mask')

Files/ processes to be updated when new w.l.-values become available (whole scheme when ISIS configuration changes)

DTM: 50m or 100m cells, vert. accuracy ±

0.5 m (?)(.adf)

Subtract DTM from focal area rasters

A-1

A-2

B-4

B-3

B-2B-1

A-7

A-6

A-5

A-4

A-3

B-5B-7

B-6B-8

(B-11)

(B-12) (B-10)(B-9)

B-15

B-17

(B-16)

C-7

C-5

B-14C-4

C-6

C-2

C-13

C-10

C-8Focal area (raster) in

Camb. with w.l.'s(.adf)

Focal area (raster) in Viet Nam with w.l.'s

(.adf)

C-9

Focal area (raster) in Camb. with water

depths(.adf)

Focal area (raster) in Viet Nam with water

depths(.adf)

ISIS reservoir polygons in Camb. with

node ID and water levels(.shp)

ISIS reservoir polygonsin Viet Nam with node

ID and water levels(.shp)

C-11

C-12

Section CCreation of

flooding maps

Section BCreation of

spatial data set

Section ATransfer of ISIS water levelsinto 'georeferenceable' table

ISIS reservoirs (raster) in Viet Nam with w.l.'s

(.adf)

ISIS reservoirs (raster) in Camb. with w.l.'s

(.adf)

Mask of focal area in Camb.(.shp)

Mask of focal area in Viet Nam

(.shp)

ISIS cross-section nodes (names,

location and w.l.'s) of Viet Nam

(.shp)

A-5: step(B-11): optional step, depending on base data set

v3a bpg fra final fmmp 20131213ok...final report, volume 3a “best practice guidelines for flood...

Documents